WEBVTT 00:00:00.000 --> 00:00:03.560 Right now, without even realizing it, your brain 00:00:03.560 --> 00:00:06.160 is actively hallucinating parallel universes. 00:00:06.879 --> 00:00:08.699 Just to help you make sense of this sentence. 00:00:08.900 --> 00:00:11.259 Yeah, it's completely invisible to us. We tend 00:00:11.259 --> 00:00:13.759 to treat human logic like a calculator, you know? 00:00:13.939 --> 00:00:16.480 Like, we just add up facts and spit out conclusions. 00:00:16.660 --> 00:00:19.440 Right, but behind the scenes, your mind is constantly 00:00:19.440 --> 00:00:22.600 taking entirely unrelated realities, stripping 00:00:22.600 --> 00:00:25.920 them for parts, and... Well, smashing them together. 00:00:26.079 --> 00:00:29.559 To create mutant ideas. And you don't even notice 00:00:29.559 --> 00:00:32.159 you're doing it. No, not at all. I mean, we treat 00:00:32.159 --> 00:00:34.759 a sudden realization or the understanding of 00:00:34.759 --> 00:00:37.140 a complex metaphor like a magic trick. The rabbit 00:00:37.140 --> 00:00:39.280 just appears out of the hat. Precisely. And we 00:00:39.280 --> 00:00:41.460 rarely stop to ask how that rabbit was actually 00:00:41.460 --> 00:00:44.259 engineered in the dark. Well, today we are going 00:00:44.259 --> 00:00:47.039 into the dark. Welcome to the deep dive. For 00:00:47.039 --> 00:00:49.920 this one, we are exploring a single beautifully 00:00:49.920 --> 00:00:53.259 dense source, a comprehensive Wikipedia article 00:00:53.259 --> 00:00:56.179 on a cognitive linguistics theory known as conceptual 00:00:56.179 --> 00:00:58.880 blending. It's a fascinating topic. Our mission 00:00:58.880 --> 00:01:01.560 today is to pull back the curtain on that invisible 00:01:01.560 --> 00:01:04.670 machinery. We're going to map out the hidden 00:01:04.670 --> 00:01:07.450 architecture of human creativity and reasoning, 00:01:07.989 --> 00:01:10.310 and hopefully give you a completely new way to 00:01:10.310 --> 00:01:12.750 look at your own mind. And what's truly fascinating 00:01:12.750 --> 00:01:15.510 here is that we aren't just examining a neat 00:01:15.510 --> 00:01:18.569 little psychological quirk. We are looking at 00:01:18.569 --> 00:01:20.870 the foundational processes of how you think. 00:01:21.030 --> 00:01:24.030 How human beings communicate. Exactly. And honestly, 00:01:24.069 --> 00:01:26.950 how we build entire civilizations. This theory 00:01:26.950 --> 00:01:30.209 attempts to decode the underlying software of 00:01:30.209 --> 00:01:32.849 everyday thought. Okay, let's unpack this. Our 00:01:32.849 --> 00:01:35.209 source refers to this theory as conceptual blending, 00:01:35.569 --> 00:01:39.390 but also conceptual integration or view application. 00:01:39.890 --> 00:01:41.890 It was formally developed by Gilles Fauconnier 00:01:41.890 --> 00:01:45.560 and Mark Turner back in 1993. Yes, though Fourconier 00:01:45.560 --> 00:01:48.519 and Turner explicitly point back to a 1964 book. 00:01:48.640 --> 00:01:50.859 Oh, right, by Arthur Koestler. Right, The Act 00:01:50.859 --> 00:01:53.739 of Creation. Koestler observed a common pattern 00:01:53.739 --> 00:01:56.260 across art, science, and humor. He called it 00:01:56.260 --> 00:01:58.500 the bisociation of matrices. Which is quite a 00:01:58.500 --> 00:02:01.319 mouthful. It is. But he essentially argued that 00:02:01.319 --> 00:02:03.900 the human mind has this unique ability to take 00:02:03.900 --> 00:02:07.340 two completely unrelated frameworks, or matrices, 00:02:07.920 --> 00:02:10.120 and suddenly snap them together to create something 00:02:10.120 --> 00:02:13.080 entirely novel. I have to admit, by association 00:02:13.080 --> 00:02:15.580 of matrices, sounds like an incredibly dry math 00:02:15.580 --> 00:02:18.240 concept. When I was reading the source, the analogy 00:02:18.240 --> 00:02:21.460 I kept coming back to was mental genetic engineering. 00:02:21.759 --> 00:02:24.560 Oh, I like that. is conceptual blending, basically 00:02:24.560 --> 00:02:27.400 taking the DNA, like the specific traits from 00:02:27.400 --> 00:02:30.159 completely different ideas, and subconsciously 00:02:30.159 --> 00:02:32.800 splicing them together in a lab to create a brand 00:02:32.800 --> 00:02:35.840 new mutant idea, one that somehow survives and 00:02:35.840 --> 00:02:38.340 makes sense to us. That is a highly accurate 00:02:38.340 --> 00:02:40.340 way to look at it. And the crucial word you use 00:02:40.340 --> 00:02:43.280 there is subconsciously. Because we don't feel 00:02:43.280 --> 00:02:46.060 it happening. Right. Foucault and Turner argued 00:02:46.060 --> 00:02:48.560 that this splicing process isn't reserved for 00:02:48.560 --> 00:02:51.460 brilliant inventors or poets. It is absolutely 00:02:51.460 --> 00:02:53.580 ubiquitous. You were doing it constantly, effortlessly, 00:02:53.800 --> 00:02:55.620 every single time you speak. Give me an example. 00:02:55.879 --> 00:02:58.620 Well, think about a common everyday phrase, like 00:02:58.620 --> 00:03:00.759 telling someone they are digging their own grave 00:03:00.759 --> 00:03:02.879 when they make a really bad financial decision. 00:03:03.050 --> 00:03:05.569 Right, because I'm not literally picturing my 00:03:05.569 --> 00:03:08.150 friend holding a shovel in a cemetery. No, of 00:03:08.150 --> 00:03:10.129 course not. I'm taking the concept of financial 00:03:10.129 --> 00:03:12.930 ruin, and I'm splicing it with the concept of 00:03:12.930 --> 00:03:16.050 physical death and burial, and my brain just 00:03:16.050 --> 00:03:18.129 instantly understands the severity of what's 00:03:18.129 --> 00:03:19.729 happening. Without actually believing your friend 00:03:19.729 --> 00:03:22.449 is a gravedigger. Exactly. You splice the DNA 00:03:22.449 --> 00:03:24.530 of those two realities together to understand 00:03:24.530 --> 00:03:28.409 a complex concept. Focconi and Turner view this 00:03:28.409 --> 00:03:31.069 theory as an attempt to create a unitary account 00:03:31.069 --> 00:03:34.289 of how cultural ideas are transmitted. So it's 00:03:34.289 --> 00:03:36.289 kind of like memetics. It's very much akin to 00:03:36.289 --> 00:03:39.409 memetics, yeah. How ideas spread, mutate, and 00:03:39.409 --> 00:03:42.560 evolve across human history. Our sources even 00:03:42.560 --> 00:03:44.900 mentioned that researchers like George Lakoff 00:03:44.900 --> 00:03:48.159 and Rafael Núñez used this exact theory to argue 00:03:48.159 --> 00:03:50.539 that the splicing is literally where mathematics 00:03:50.539 --> 00:03:53.439 comes from. Yes, the cognitive science of mathematics. 00:03:53.580 --> 00:03:55.860 Now that understanding math isn't just about 00:03:55.860 --> 00:03:58.280 recognizing universal truths floating in space. 00:03:58.699 --> 00:04:01.699 It actually requires the mastery of these extensive 00:04:01.699 --> 00:04:05.240 metaphorical blends. Precisely. Mathematics isn't 00:04:05.240 --> 00:04:07.460 just out there in the ether waiting to be discovered. 00:04:07.610 --> 00:04:10.389 According to this view, it is built by human 00:04:10.389 --> 00:04:13.830 minds taking basic grounded concepts like physical 00:04:13.830 --> 00:04:16.889 objects. Yeah, physical objects or grouping things 00:04:16.889 --> 00:04:19.889 together in piles and then blending them over 00:04:19.889 --> 00:04:22.889 and over again into increasingly abstract mental 00:04:22.889 --> 00:04:25.430 spaces. I want to take this from the abstract 00:04:25.430 --> 00:04:28.370 down into a concrete real -world application 00:04:28.370 --> 00:04:30.670 because I want you, the listener, to actually 00:04:30.670 --> 00:04:34.310 feel your brain doing this mental genetic splicing 00:04:34.310 --> 00:04:36.870 in real time. It's one thing to talk about it, 00:04:36.990 --> 00:04:39.490 but experiencing it is different. Right. To do 00:04:39.490 --> 00:04:41.750 that, we are going to use Arthur Koestler's famous 00:04:41.750 --> 00:04:43.949 riddle of the Buddhist monk. It's heavily featured 00:04:43.949 --> 00:04:45.970 in the sources. Oh, it's one of the classic examples 00:04:45.970 --> 00:04:48.610 used to demonstrate a blend. It perfectly forces 00:04:48.610 --> 00:04:50.569 your brain to build the machinery we're talking 00:04:50.569 --> 00:04:52.970 about. OK, so here is the middle. Picture this. 00:04:53.529 --> 00:04:55.889 A Buddhist monk begins at dawn one day walking 00:04:55.889 --> 00:04:58.519 up a mountain path. He reaches the top at sunset. 00:04:58.819 --> 00:05:00.959 Okay. He meditates at the top for several days. 00:05:01.480 --> 00:05:04.199 Then, at dawn on a new day, he begins to walk 00:05:04.199 --> 00:05:06.660 back down that exact same mountain path, reaching 00:05:06.660 --> 00:05:08.939 the foot of the mountain at sunset. And here 00:05:08.939 --> 00:05:11.120 are the constraints for you listening. You can 00:05:11.120 --> 00:05:13.600 make absolutely no assumptions about his starting 00:05:13.600 --> 00:05:17.019 or stopping or about his pace during either trip. 00:05:17.240 --> 00:05:19.040 Right, you could sprint for an hour, he could 00:05:19.040 --> 00:05:21.259 crawl. He could take a three -hour nap on a rock. 00:05:21.860 --> 00:05:24.480 We don't know. The riddle is this. prove that 00:05:24.480 --> 00:05:26.519 there is a single place on the path that the 00:05:26.519 --> 00:05:29.300 monk occupies at the exact same hour of the day 00:05:29.300 --> 00:05:31.800 on both of those separate journeys. It's a tricky 00:05:31.800 --> 00:05:34.620 one. When I first read this, my brain immediately 00:05:34.620 --> 00:05:37.259 tried to draw a graph. I was trying to calculate 00:05:37.259 --> 00:05:41.240 velocity and time, and it was incredibly frustrating. 00:05:41.420 --> 00:05:44.000 I'm sure. The variables of his speed and his 00:05:44.000 --> 00:05:46.519 breaks make a mathematical approach feel completely 00:05:46.519 --> 00:05:48.839 impossible. It does feel impossible if you rely 00:05:48.839 --> 00:05:52.990 strictly on monotonic linear logic. but your 00:05:52.990 --> 00:05:55.410 brain has a shortcut. How did you find it? Well, 00:05:55.490 --> 00:05:57.910 the only way to solve it easily is to abandon 00:05:57.910 --> 00:06:00.930 reality. You have to imagine a completely fictional, 00:06:01.329 --> 00:06:03.589 improbable scenario. Go on. You have to take 00:06:03.589 --> 00:06:06.050 day one and day two and superimpose them. You 00:06:06.050 --> 00:06:08.790 just picture two separate monks. One is at the 00:06:08.790 --> 00:06:10.850 bottom of the mountain walking up. The other 00:06:10.850 --> 00:06:13.490 is at the top of the mountain walking down. They 00:06:13.490 --> 00:06:16.449 both start at dawn on the exact same day. And 00:06:16.449 --> 00:06:18.290 what happens in your mind's eye when you watch 00:06:18.290 --> 00:06:21.269 that play out? Oh, I see it immediately. If they 00:06:21.269 --> 00:06:23.550 are on the exact same path, one going up and 00:06:23.550 --> 00:06:26.529 one going down on the same day, they physically 00:06:26.529 --> 00:06:28.750 have to collide at some point. They will cross 00:06:28.750 --> 00:06:31.209 paths. And where they meet, that's the proof. 00:06:31.329 --> 00:06:33.449 That is the spot on the path they occupy at the 00:06:33.449 --> 00:06:37.050 exact same hour. It is so elegantly simple once 00:06:37.050 --> 00:06:39.430 you see it. It is. But what really strikes me 00:06:39.430 --> 00:06:43.459 here isn't the math. It's that To solve a completely 00:06:43.459 --> 00:06:47.360 factual, logical problem, my brain naturally 00:06:47.360 --> 00:06:50.339 generated a completely fictional world. You built 00:06:50.339 --> 00:06:52.420 a new reality. We preserved the time of day. 00:06:52.480 --> 00:06:55.259 We preserved the physical path. But we literally 00:06:55.259 --> 00:06:59.139 cloned the monk. We hallucinated a parallel universe 00:06:59.139 --> 00:07:01.759 with two monks just to make the logic work. And 00:07:01.759 --> 00:07:03.639 you did it instantly. You didn't consciously 00:07:03.639 --> 00:07:05.579 sit down and instruct your brain, OK, let us 00:07:05.579 --> 00:07:08.360 now construct a parallel universe. You just blended 00:07:08.360 --> 00:07:10.860 the scenarios. Right. So now that we've experienced 00:07:10.860 --> 00:07:14.120 a firsthand, we need to dissect how the brain 00:07:14.120 --> 00:07:16.959 just pulled it off. Yeah, if I picture my working 00:07:16.959 --> 00:07:20.089 memory like an animation studio... What are the 00:07:20.089 --> 00:07:22.670 actual layers I'm putting together to make that 00:07:22.670 --> 00:07:25.329 final scene? The source is called these mental 00:07:25.329 --> 00:07:29.170 spaces. Yes. Think of mental spaces as transparent 00:07:29.170 --> 00:07:32.050 layers of animation cells being stacked on top 00:07:32.050 --> 00:07:34.269 of each other. Okay, transparent layers. They 00:07:34.269 --> 00:07:36.750 are small conceptual containers generated right 00:07:36.750 --> 00:07:39.550 there in your working memory, but they pull knowledge 00:07:39.550 --> 00:07:43.170 from your vast long -term memory. In a basic 00:07:43.170 --> 00:07:45.310 integration network, there are at least four 00:07:45.310 --> 00:07:47.509 of these interconnected transparent layers stacked 00:07:47.509 --> 00:07:50.500 up. Four separate mental containers just to solve 00:07:50.500 --> 00:07:52.899 the monk riddle. Let's reason through them. First, 00:07:53.040 --> 00:07:55.160 you have the generic space. Right. The generic 00:07:55.160 --> 00:07:57.519 space captures only the common structure shared 00:07:57.519 --> 00:08:00.339 by your inputs. It's the barest of bones. So 00:08:00.339 --> 00:08:03.199 no details. Exactly. In the riddle, the generic 00:08:03.199 --> 00:08:05.879 space contains a mountain path, the abstract 00:08:05.879 --> 00:08:08.379 concept of a day passing, and a person moving 00:08:08.379 --> 00:08:11.019 along that path. Nothing specific. Just the rules 00:08:11.019 --> 00:08:13.279 of the world. Then you need the actual material. 00:08:13.339 --> 00:08:15.959 That's the two input spaces. Yeah. So input space 00:08:15.959 --> 00:08:19.199 one is my first animation cell. Day one, the 00:08:19.199 --> 00:08:21.660 monk walking up the mountain. Yes. Input space 00:08:21.660 --> 00:08:24.540 two is my second cell. Day two, the monk walking 00:08:24.540 --> 00:08:27.120 down the mountain. Exactly. And then the integration 00:08:27.120 --> 00:08:28.959 happens in the fourth container, which is the 00:08:28.959 --> 00:08:30.980 blended space. The final stack. This is where 00:08:30.980 --> 00:08:33.720 all the transparent cells stack together. The 00:08:33.720 --> 00:08:35.659 blended space takes the general structure from 00:08:35.659 --> 00:08:38.259 the generic space and through a process called 00:08:38.259 --> 00:08:41.230 selective projection. It pulls specific elements 00:08:41.230 --> 00:08:43.750 from input one and input two. Wait, selective 00:08:43.750 --> 00:08:45.929 projection. That means I'm not just mashing everything 00:08:45.929 --> 00:08:48.129 together thoughtlessly. Right. Notice the word 00:08:48.129 --> 00:08:51.850 selective. We keep the mountain path as one single 00:08:51.850 --> 00:08:54.129 element. We keep the time of day as one single 00:08:54.129 --> 00:08:56.289 element. We don't have two mountains or a 48 00:08:56.289 --> 00:08:58.450 -hour day. Oh, I see. But we project the monk 00:08:58.450 --> 00:09:01.769 twice. Because the motion's direction is different 00:09:01.769 --> 00:09:04.710 in each input, your brain selectively decides 00:09:04.710 --> 00:09:07.409 to create two separate monks in the blended space 00:09:07.409 --> 00:09:10.139 to represent those two directions. But how does 00:09:10.139 --> 00:09:12.419 the new reality actually start moving? I've got 00:09:12.419 --> 00:09:15.419 my animation cells stacked up. Two monks, one 00:09:15.419 --> 00:09:18.779 path, one day. How do the monks suddenly know 00:09:18.779 --> 00:09:20.820 they're supposed to walk toward each other and 00:09:20.820 --> 00:09:22.919 meet? That's where the three operations of blending 00:09:22.919 --> 00:09:25.500 come in. This results in what the theory calls 00:09:25.500 --> 00:09:28.600 an emergent structure. An emergent structure. 00:09:28.700 --> 00:09:31.200 Yes. This is a structure or a realization that 00:09:31.200 --> 00:09:33.559 did not exist in either of the original inputs. 00:09:33.740 --> 00:09:36.279 So the meaning of the monks is the emergent structure. 00:09:36.440 --> 00:09:38.649 Exactly. The first operation to get there is 00:09:38.649 --> 00:09:41.309 composition. This is simply the act of bringing 00:09:41.309 --> 00:09:44.090 the elements together onto the same cell. You 00:09:44.090 --> 00:09:46.269 compose the scene by putting both monks on the 00:09:46.269 --> 00:09:48.909 same path on the same day. Setting the stage. 00:09:49.190 --> 00:09:52.309 Got it. What's next? Second is completion. This 00:09:52.309 --> 00:09:54.370 is where your brain reaches into its long -term 00:09:54.370 --> 00:09:56.950 memory and passes on additional background meaning 00:09:56.950 --> 00:09:59.409 to the blend. Like the laws of physics. Exactly. 00:09:59.659 --> 00:10:02.320 Your brain already knows how physical space works. 00:10:02.799 --> 00:10:04.899 It knows that two solid objects moving toward 00:10:04.899 --> 00:10:07.559 each other on a single path will eventually collide. 00:10:08.299 --> 00:10:09.960 You don't have to relearn physics every time 00:10:09.960 --> 00:10:12.659 you think. Your brain automatically completes 00:10:12.659 --> 00:10:15.419 the logic of the scene. And the third operation. 00:10:15.720 --> 00:10:18.379 Elaboration. This is the dynamic running of the 00:10:18.379 --> 00:10:22.080 blend. You press play on the animation. Ah, okay. 00:10:22.159 --> 00:10:24.440 You watch the two monks walk in your mind's eye 00:10:24.440 --> 00:10:26.960 and you watch them meet. The meeting only exists 00:10:26.960 --> 00:10:29.659 in the blend. but it solves your real -world 00:10:29.659 --> 00:10:31.639 problem. It's incredible when you break it down 00:10:31.639 --> 00:10:34.399 into mechanics like that. But the sources are 00:10:34.399 --> 00:10:36.759 quick to point out that this monk riddle is just 00:10:36.759 --> 00:10:40.639 one specific type of blend. The theory categorizes 00:10:40.639 --> 00:10:43.740 the vast chaos of human thought into four main 00:10:43.740 --> 00:10:46.360 types of integration network. Oh yes, and moving 00:10:46.360 --> 00:10:49.000 through these four networks really shows how 00:10:49.000 --> 00:10:52.440 scalable this theory is. The Buddhist monk riddle 00:10:52.440 --> 00:10:54.740 we just did is an example of a mirror network. 00:10:54.919 --> 00:10:57.379 Because the two inputs mirror each other. Because 00:10:57.379 --> 00:10:59.879 there is a shared organizing frame present in 00:10:59.879 --> 00:11:02.700 all the mental spaces. In all the spaces, you 00:11:02.700 --> 00:11:04.659 have a man making a journey on a mountain path. 00:11:05.080 --> 00:11:08.740 The core activity mirrors itself. But not all 00:11:08.740 --> 00:11:10.700 blends are that symmetrical. Right. The sources 00:11:10.700 --> 00:11:12.480 list three others. Let's try to ground these. 00:11:12.539 --> 00:11:15.059 The most basic one is the simplex network. A 00:11:15.059 --> 00:11:17.500 simplex network is essentially just dropping 00:11:17.500 --> 00:11:20.019 values into a predefined frame. Like filling 00:11:20.019 --> 00:11:22.899 in a blank. Basically, if I say Paul is the father 00:11:22.899 --> 00:11:25.399 of Sally, You have a frame for family roles, 00:11:26.000 --> 00:11:28.000 a father and a daughter. That's one input. The 00:11:28.000 --> 00:11:30.139 other input is just the specific people, Paul 00:11:30.139 --> 00:11:32.220 and Sally. And you blend them together? And you 00:11:32.220 --> 00:11:35.019 instantly understand their relationship. No parallel 00:11:35.019 --> 00:11:37.600 universes required. Okay, simple enough. Then 00:11:37.600 --> 00:11:40.139 there's the single scope network. The sources 00:11:40.139 --> 00:11:42.179 say this is where you have two totally different 00:11:42.179 --> 00:11:45.139 frames, but only one gets projected into the 00:11:45.139 --> 00:11:47.639 blend to organize it. A perfect example of a 00:11:47.639 --> 00:11:50.480 single -scope blend is how you understand a computer 00:11:50.480 --> 00:11:52.840 desktop. Oh, interesting. You have an input space 00:11:52.840 --> 00:11:56.200 for a physical office, environment desks, folders, 00:11:56.500 --> 00:11:59.100 trash cans. You have another input space for 00:11:59.100 --> 00:12:01.539 digital computer code and file directories. Oh, 00:12:01.539 --> 00:12:04.570 wow. I see it. We don't organize our screens 00:12:04.570 --> 00:12:07.730 by reading raw code. We project the organizing 00:12:07.730 --> 00:12:10.710 frame of the physical office onto the digital 00:12:10.710 --> 00:12:14.370 space. Yes. We literally drag a digital file 00:12:14.370 --> 00:12:18.230 into a digital trash can. It's two completely 00:12:18.230 --> 00:12:20.389 different realities, but the physical office 00:12:20.389 --> 00:12:22.990 frame organizes the blended space. Precisely. 00:12:24.090 --> 00:12:26.970 And that brings us to the most complex one, the 00:12:26.970 --> 00:12:29.799 double -scope network. This is where you pull 00:12:29.799 --> 00:12:33.259 parts of the organizing frames from both inputs 00:12:33.259 --> 00:12:36.259 to create something really wild, often resulting 00:12:36.259 --> 00:12:38.440 in frame clashes. This would be like a computer 00:12:38.440 --> 00:12:41.440 virus, right? Yes. You have the biological frame 00:12:41.440 --> 00:12:44.360 of a virus, an organism that infects, replicates, 00:12:44.419 --> 00:12:46.860 and destroys a host, and you have the computer 00:12:46.860 --> 00:12:50.299 frame of software code. Right. You pull the replication 00:12:50.299 --> 00:12:53.460 and infection from biology and you pull the digital 00:12:53.460 --> 00:12:55.799 environment from the computer and you create 00:12:55.799 --> 00:12:58.799 a mutant idea. malicious code that acts like 00:12:58.799 --> 00:13:01.200 a biological pathogen. Exactly. Double -scope 00:13:01.200 --> 00:13:03.500 networks are the engine of high -level human 00:13:03.500 --> 00:13:06.299 creativity. And navigating all these different 00:13:06.299 --> 00:13:08.179 networks, whether it's a computer virus or the 00:13:08.179 --> 00:13:10.519 monk riddle, requires us to manipulate what the 00:13:10.519 --> 00:13:13.100 theory calls vital relations. Okay, I really 00:13:13.100 --> 00:13:15.000 want to push back on this idea of vital relations 00:13:15.000 --> 00:13:16.820 because it seems like the cheat code that makes 00:13:16.820 --> 00:13:19.600 the whole theory work. In the monk riddle, time 00:13:19.600 --> 00:13:22.580 was completely manipulated. We compressed two 00:13:22.580 --> 00:13:24.980 separate days into a single day so the monk could 00:13:24.980 --> 00:13:27.799 walk up and down simultaneously. But if we are 00:13:27.799 --> 00:13:30.700 constantly blending realities, we must be manipulating 00:13:30.700 --> 00:13:34.230 other things besides just time. Right. What other 00:13:34.230 --> 00:13:37.110 rules of reality are we subconsciously bending 00:13:37.110 --> 00:13:39.529 to our will? Well, time compression is the most 00:13:39.529 --> 00:13:42.250 obvious, but the sources list several vital relations 00:13:42.250 --> 00:13:44.970 we manipulate constantly. We manipulate cause 00:13:44.970 --> 00:13:48.769 effect. We manipulate change, how an entity transforms 00:13:48.769 --> 00:13:52.309 over time. We manipulate space, compressing vast 00:13:52.309 --> 00:13:54.590 geographic distances so we can compare things 00:13:54.590 --> 00:13:57.190 side by side in our minds. Like saying New York 00:13:57.190 --> 00:13:59.590 is breathing down London's neck in a financial 00:13:59.590 --> 00:14:02.169 context. We compress an ocean of space to create 00:14:02.169 --> 00:14:05.210 a physical race. Yes. Perfect example. We also 00:14:05.210 --> 00:14:07.970 manipulate identity and role. Think about when 00:14:07.970 --> 00:14:09.669 someone says, if I were you, I would quit that 00:14:09.669 --> 00:14:12.009 job. Right. You are blending your identity with 00:14:12.009 --> 00:14:14.110 their role. You maintain your own decision -making 00:14:14.110 --> 00:14:16.509 logic, but you project it into their life circumstances. 00:14:16.590 --> 00:14:19.299 Oh, I do that all the time. We all do. These 00:14:19.299 --> 00:14:21.799 vital relations are the threads that allow us 00:14:21.799 --> 00:14:24.960 to stitch different mental spaces together without 00:14:24.960 --> 00:14:27.779 our brains glitching out. And speaking of glitching 00:14:27.779 --> 00:14:30.179 out, if we've managed to map out human thought 00:14:30.179 --> 00:14:33.899 so systematically, literally breaking it down 00:14:33.899 --> 00:14:37.419 into four interconnected mental containers, three 00:14:37.419 --> 00:14:40.559 operations, and specific vital relations, it 00:14:40.559 --> 00:14:43.419 raises a massive obvious question. I think I 00:14:43.419 --> 00:14:44.820 know where you're going with this. Here's where 00:14:44.820 --> 00:14:47.419 it gets really interesting. If we have the blueprint, 00:14:47.639 --> 00:14:49.539 why can't we just code this into a computer? 00:14:50.120 --> 00:14:53.419 The artificial intelligence question. If conceptual 00:14:53.419 --> 00:14:55.820 blending is truly the architecture of thought, 00:14:56.299 --> 00:14:58.419 can we build an artificial brain with the same 00:14:58.419 --> 00:15:00.899 architecture? Our sources actually dive deep 00:15:00.899 --> 00:15:03.480 into this. The attempt to model this mathematically 00:15:03.480 --> 00:15:06.330 isn't new at all. Even before conceptual blending 00:15:06.330 --> 00:15:08.649 was fully formalized by Fauconnier and Turner, 00:15:09.230 --> 00:15:11.690 an early computational model of a closely related 00:15:11.690 --> 00:15:14.110 process called view application was implemented 00:15:14.110 --> 00:15:16.830 in the 1980s. Yes, by a researcher named Jeff 00:15:16.830 --> 00:15:18.990 Schrager. Right, at Carnegie Mellon University. 00:15:19.370 --> 00:15:21.769 He applied it to causal reasoning about complex 00:15:21.769 --> 00:15:24.940 devices. But modern AI has struggled profoundly 00:15:24.940 --> 00:15:27.799 with this. They are trying to build what sources 00:15:27.799 --> 00:15:31.759 call non -monotonic reasoning into AI systems 00:15:31.759 --> 00:15:35.240 to handle complex human -like concept combinations. 00:15:35.639 --> 00:15:38.259 Okay, let's pause there because non -monotonic 00:15:38.259 --> 00:15:41.700 reasoning sounds incredibly dense. From my understanding, 00:15:41.960 --> 00:15:44.179 normal machine logic is monotonic. It builds 00:15:44.179 --> 00:15:46.679 up in a straight, irreversible line. Yes. If 00:15:46.679 --> 00:15:48.980 A is true and B is true, they just add together. 00:15:49.519 --> 00:15:52.840 But human logic is non -monotonic, meaning we 00:15:52.840 --> 00:15:55.139 can learn a new fact that completely contradicts 00:15:55.139 --> 00:15:57.720 and erases an old conclusion. We can rewrite 00:15:57.720 --> 00:15:59.899 the rules on the fly. That is a brilliant way 00:15:59.899 --> 00:16:02.919 to explain it. In monotonic logic, adding new 00:16:02.919 --> 00:16:05.860 facts only builds up a conclusion. In non -monotonic 00:16:05.860 --> 00:16:08.299 logic, a new piece of information might force 00:16:08.299 --> 00:16:10.320 you to completely withdraw a previous assumption. 00:16:11.139 --> 00:16:13.039 And the classic example cited in our sources 00:16:13.039 --> 00:16:16.100 for this computational hurdle in AI is the famous 00:16:16.100 --> 00:16:18.759 pet fish problem. I love the pet fish problem 00:16:18.759 --> 00:16:21.039 because it perfectly illustrates the gap between 00:16:21.039 --> 00:16:23.480 human blending and machine logic. It really does. 00:16:23.659 --> 00:16:26.159 If you tell an AI about a pet, it activates a 00:16:26.159 --> 00:16:28.600 frame. Pets are furry, they live in houses, they're 00:16:28.600 --> 00:16:30.440 affectionate, they seek attention like a dog 00:16:30.440 --> 00:16:33.379 or a cat. Right. And if you tell that same AI 00:16:33.379 --> 00:16:35.860 about a fish, it activates a completely different 00:16:35.860 --> 00:16:38.590 frame. Fish live in the ocean. They are scaly. 00:16:38.789 --> 00:16:40.629 They are cold -blooded. They're often caught 00:16:40.629 --> 00:16:42.929 for food, like a salmon or a shark. But if you 00:16:42.929 --> 00:16:45.690 say the words pet fish together, a human being 00:16:45.690 --> 00:16:49.350 instantly creates a blended space. We don't picture 00:16:49.350 --> 00:16:52.450 a furry, affectionate salmon living on our living 00:16:52.450 --> 00:16:55.549 room couch. And we don't picture a cuddly great 00:16:55.549 --> 00:16:58.309 white shark. No, we don't. We instantly generate 00:16:58.309 --> 00:17:01.049 an emergent structure. A small goldfish living 00:17:01.049 --> 00:17:03.450 in a glass bowl. And we effortlessly discard 00:17:03.450 --> 00:17:06.180 the irrelevant parts of the pet frame. the fur, 00:17:06.420 --> 00:17:09.039 the cuddling, and we discard the irrelevant parts 00:17:09.039 --> 00:17:11.900 of the fish frame, the ocean, the food aspect. 00:17:12.220 --> 00:17:14.960 We selectively project only what works. But an 00:17:14.960 --> 00:17:17.220 AI naturally wants to strictly add the two frames 00:17:17.220 --> 00:17:20.319 together. It wants monotonic logic. A pet is 00:17:20.319 --> 00:17:22.880 furry plus a fish is a fish equals a furry fit. 00:17:23.279 --> 00:17:25.779 Exactly. Teaching a machine to selectively project 00:17:25.779 --> 00:17:28.140 and create an emergent structure that is completely 00:17:28.140 --> 00:17:31.119 different from its original inputs is computationally 00:17:31.119 --> 00:17:34.380 agonizing. It requires immense creativity frameworks. 00:17:34.680 --> 00:17:37.819 And honestly, the fact that a supercomputer crashes 00:17:37.819 --> 00:17:40.460 trying to imagine a goldfish in a ball brings 00:17:40.460 --> 00:17:43.319 up a glaring issue with this whole theory. It 00:17:43.319 --> 00:17:45.859 acts as a perfect transition into the philosophical 00:17:45.859 --> 00:17:48.720 pushbacks. Yes, the critiques. If conceptual 00:17:48.720 --> 00:17:51.660 blending is so hard to program, is it actually 00:17:51.660 --> 00:17:54.660 a rigid predictive science, or is it merely a 00:17:54.660 --> 00:17:57.079 descriptive framework? Let's bring in the critics, 00:17:57.240 --> 00:17:59.079 because they have a lot to say. They certainly 00:17:59.079 --> 00:18:02.200 do. The main skepticism comes from Raymond W. 00:18:02.339 --> 00:18:04.440 Gibbs Jr., who published a heavy critique in 00:18:04.440 --> 00:18:08.140 2000. Gibbs essentially points out a major flaw 00:18:08.140 --> 00:18:10.880 that you, the listener, might already be sensing. 00:18:11.579 --> 00:18:14.619 The theory lacks testable hypotheses. Which is 00:18:14.619 --> 00:18:17.660 a big problem. If a psychological theory is going 00:18:17.660 --> 00:18:20.480 to be treated as hard science, it needs to be 00:18:20.480 --> 00:18:23.700 able to predict behavior. Gibbs argued that blending 00:18:23.700 --> 00:18:25.880 theory is really just a framework. It's a way 00:18:25.880 --> 00:18:28.539 of describing things after the fact, not a single 00:18:28.539 --> 00:18:30.480 testable theory that predicts what someone will 00:18:30.480 --> 00:18:32.859 think next. The analogy that came to mind when 00:18:32.859 --> 00:18:35.500 I read Gibbs's critique was baking a cake. How 00:18:35.500 --> 00:18:38.670 so? Gibbs is essentially arguing that inferring 00:18:38.670 --> 00:18:41.049 how a mental process works just by looking at 00:18:41.049 --> 00:18:44.150 the final product is deeply flawed. It's like 00:18:44.150 --> 00:18:46.750 looking at a fully baked chocolate cake and claiming 00:18:46.750 --> 00:18:50.230 you know exactly, with 100 % certainty, the exact 00:18:50.230 --> 00:18:52.430 order the baker added the eggs, the flour, and 00:18:52.430 --> 00:18:54.349 the sugar. That's a great way to put it. We can 00:18:54.349 --> 00:18:56.970 look at the final blend in your head, the goldfish 00:18:56.970 --> 00:18:59.670 or the monk, but we can't actually prove your 00:18:59.670 --> 00:19:02.609 brain used four separate distinct mental containers 00:19:02.609 --> 00:19:05.450 to get there. It's a devastating critique. We 00:19:05.450 --> 00:19:07.849 can guess the recipe, but we don't really know 00:19:07.849 --> 00:19:10.250 what happened in the bowl. Gibbs even suggested 00:19:10.250 --> 00:19:12.349 that other linguistic theories might be equally 00:19:12.349 --> 00:19:15.029 effective at explaining these cognitive phenomena 00:19:15.029 --> 00:19:18.190 without requiring this incredibly specific architecture 00:19:18.190 --> 00:19:20.619 of mental spaces. And I want to jump on another 00:19:20.619 --> 00:19:23.400 criticism here from David Ritchie in 2004, who 00:19:23.400 --> 00:19:26.079 argued against the unnecessary complexity of 00:19:26.079 --> 00:19:28.000 the theory. Right, the Occam's razor approach. 00:19:28.240 --> 00:19:30.700 Exactly. Ritchie brings Occam's razor to the 00:19:30.700 --> 00:19:33.700 table. He asks, does every single metaphor we 00:19:33.700 --> 00:19:36.680 use really require this massive invisible machinery? 00:19:37.240 --> 00:19:39.279 Think back to our earlier example of digging 00:19:39.279 --> 00:19:41.960 your own grave or just basic trash talk. Yeah. 00:19:41.980 --> 00:19:44.900 Are we really generating a generic space, two 00:19:44.900 --> 00:19:47.940 input spaces and a blended space running three 00:19:47.940 --> 00:19:50.660 distinct operations? and manipulating vital relations 00:19:50.660 --> 00:19:53.759 just to understand a simple insult. Ritchie argues 00:19:53.759 --> 00:19:56.619 that it is highly possible we are overcomplicating 00:19:56.619 --> 00:19:59.500 what might be a much simpler, faster cognitive 00:19:59.500 --> 00:20:02.559 process. Even with the Buddhist monk riddle, 00:20:02.880 --> 00:20:05.259 Ritchie argues there are alternative, simpler 00:20:05.259 --> 00:20:07.960 interpretations for how human beings solve it 00:20:07.960 --> 00:20:10.019 without needing the complex mirror network model. 00:20:10.250 --> 00:20:12.410 But what I find most telling in all of our sources 00:20:12.410 --> 00:20:15.069 is the philosophical boundary drawn by Mark Turner 00:20:15.069 --> 00:20:18.269 himself, one of the literal creators of the theory. 00:20:18.430 --> 00:20:21.309 Yes. In his book, The Literary Mind, Turner makes 00:20:21.309 --> 00:20:23.549 a really fascinating admission that seems to 00:20:23.549 --> 00:20:25.869 agree with the critics on some level. He does. 00:20:26.250 --> 00:20:28.690 Turner states that conceptual blending is undoubtedly 00:20:28.690 --> 00:20:30.849 a fundamental instrument of the everyday mind. 00:20:31.430 --> 00:20:33.950 But he clarifies that the insights obtained from 00:20:33.950 --> 00:20:37.069 it, the metaphors, the solutions, are the products 00:20:37.069 --> 00:20:39.740 of creative thinking. The products. Yes, and 00:20:39.740 --> 00:20:42.799 this is the key limitation. He admits that conceptual 00:20:42.799 --> 00:20:45.599 blending theory is not in itself a complete theory 00:20:45.599 --> 00:20:48.819 of creativity. Because it can't explain inspiration. 00:20:49.940 --> 00:20:52.500 Turner admits that blending provides the terminology 00:20:52.500 --> 00:20:55.339 for describing how we combine realities, but 00:20:55.339 --> 00:20:58.119 it has absolutely nothing to say about where 00:20:58.119 --> 00:21:00.299 the inputs to a blend originate in the first 00:21:00.299 --> 00:21:03.000 place. It explains the mechanics of the mashup. 00:21:03.299 --> 00:21:06.579 It explains how the DNA is spliced. but it doesn't 00:21:06.579 --> 00:21:09.119 explain where the DNA came from. So what does 00:21:09.119 --> 00:21:12.200 this all mean for us? We've journeyed from the 00:21:12.200 --> 00:21:14.799 spark of an idea through the bisociation of matrices. 00:21:14.980 --> 00:21:17.079 We have. We've walked up and down the mountain 00:21:17.079 --> 00:21:20.039 with our cloned Buddhist monk. We've watched 00:21:20.039 --> 00:21:22.279 artificial intelligence struggle to understand 00:21:22.279 --> 00:21:25.319 a goldfish. And we've waded into the philosophical 00:21:25.319 --> 00:21:27.259 critiques of whether this is hard science or 00:21:27.259 --> 00:21:29.660 just a really good metaphor for thought. It means 00:21:29.660 --> 00:21:31.519 that whether it actually takes four distinct 00:21:31.519 --> 00:21:34.000 mental animation cells, or whether it's a much 00:21:34.000 --> 00:21:36.900 simpler process as Occam's razor suggests, the 00:21:36.900 --> 00:21:38.819 undeniable truth is that your brain is constantly 00:21:38.819 --> 00:21:41.000 performing high -level conceptual gymnastics. 00:21:41.180 --> 00:21:43.319 Every single day. You are continuously pulling 00:21:43.319 --> 00:21:45.640 frames from your memory, selectively projecting 00:21:45.640 --> 00:21:48.259 elements and running dynamic simulations just 00:21:48.259 --> 00:21:50.619 to make sense of a simple conversation or solve 00:21:50.619 --> 00:21:53.750 a daily problem at work. You are quite literally 00:21:53.750 --> 00:21:57.430 a walking talking engine of conceptual integration. 00:21:57.630 --> 00:21:59.250 But I would like to leave you with a final thought 00:21:59.250 --> 00:22:02.029 to mull over. Lay it on us. If conceptual blending 00:22:02.029 --> 00:22:04.490 only provides the terminology for how we mix 00:22:04.490 --> 00:22:07.069 and match existing inputs, if it only explains 00:22:07.069 --> 00:22:10.009 the recipe but not the ingredients, then where 00:22:10.009 --> 00:22:12.410 does the very first spark of a truly original 00:22:12.410 --> 00:22:15.730 idea actually come from before it is ever blended? 00:22:16.289 --> 00:22:19.190 What generates the input before the input? Wow. 00:22:19.579 --> 00:22:22.140 That is a deeply haunting question to end on. 00:22:22.380 --> 00:22:24.559 Where does the original thought come from? Well, 00:22:24.779 --> 00:22:26.420 thank you so much for joining us on this deep 00:22:26.420 --> 00:22:28.819 dive. As you go about your day today, pay attention 00:22:28.819 --> 00:22:31.180 to your own thoughts. Notice the hidden blends 00:22:31.180 --> 00:22:33.660 in your conversations, the metaphors you use 00:22:33.660 --> 00:22:35.839 without thinking, and the fictional realities 00:22:35.839 --> 00:22:38.180 you construct just to navigate the factual world. 00:22:38.599 --> 00:22:40.539 Keep questioning the invisible machinery. We'll 00:22:40.539 --> 00:22:41.140 catch you next time.