"Every time this child was proven wrong, he learned something new." In this episode, Andrew Stotz continues his conversation with Balaji Reddie as they explore two essential elements of Dr. Deming's System of Profound Knowledge: the Theory of Knowledge and Psychology. Through the memorable story of a young boy trying to understand how his father's scooter starts, Balaji illustrates how we build theories, test them through experience, and improve them through learning. The discussion also explores why PDSA begins with theory, why experience alone is not enough, and why understanding how people learn is fundamental to leading improvement. Whether you're new to Deming or deep into his work, this episode offers practical insights into how better questions lead to better knowledge and, ultimately, better management. Are you ready to learn more about the System of Profound Knowledge? Check out DemingNEXT online learning! https://deming.org/learn/demingnext/ TRANSCRIPT 0:00:02.2 Andrew Stotz: My name is Andrew Stotz, and I'll be your host as we dive deeper into the teachings of Dr. W. Edwards Deming. Today, I'm continuing my discussion with Balaji Reddie, an educator and trainer in the teachings of Dr. Deming, and quality management generally. The topic for today is: The theory of knowledge and psychology. Take it away. 0:00:29.8 Balaji Reddie: Well, thank you. Good morning, Andrew. And it's great to continue our series. So the last time we met, we had begun with the Profound Knowledge, where we said that Dr. Deming brought together four, seemingly disconnected, sciences. He never invented any single one of them, but he saw the connections, and he created, what you could say is, a bootstrap theory of management, if I may say so, a very all-encompassing view of things happening around you. That's why he had no other word to describe it except profound. Because you see things that you normally would not see. That's what we said the last time. And when he proposed that, all of a sudden, I told you the funny thing was that experts in each of those sciences jumped in. I think they realized that they were caught napping, that they did not see the connections, and they did not see a bare minimum set of competencies that you need to have in order to understand things in their entirety. It was very, very... It was amazing that he actually could put that all together. 0:01:44.3 Balaji Reddie: So we started off with the four sciences, in no order of preference. We said here... His words were appreciation for a system. That means, taking into consideration the fact that everything... Having a holistic view, that what we see is not an event but an eventuality, that there's a huge big system behind it. Whatever we get to see in front of our eyes, we're just getting to see the effect. So appreciating that, that's what he said, for a system, a systemic view. I just used a different word, I said connectedness. And I said that from now on, we say that there are no events, there are only eventualities. Which means when something happens in front of our eyes, we always question... We bring in the concept of time and space; that the origins of this, what I'm seeing in front of my eyes, lie somewhere else and maybe actually in a different time frame. So I need to study deeper in order to understand why I'm seeing this in front of my eyes. And you gave that fantastic example of coffee beans, where you said everything has to happen at the right time. So that was the first one. 0:02:54.1 Balaji Reddie: Second one, was about understanding variation. And I would say, that's statistical thinking. I think around 1998, I saw someone come out with the definition of statistical thinking. And they said that it's a holistic view of looking at a process, that everything in the world, happens in the form of a process, and all processes are subject to variation. And understanding and reducing that variation, is synonymous to increasing quality. Now, I'd just like to make a slight change there. Yes, all processes are subject to variation, but talking about reducing... I think, understanding and managing that variation, would lead to better quality. And his entire insistence was understanding variation. Although, let me tell you this, in Henry Neave's book, he wrote about this, where he said that in one of the weekend chats that they had with Deming, he made a statement that if I had to reduce my message to managers in two words, I would say: reduce variation. And this, I've heard from both Henry Neave and David Kerridge. For those who are unaware of who David Kerridge was, because he passed away, I think, around 2018 or 2019, I can't remember... Or maybe before. I'm forgetting the year. 0:04:32.1 Balaji Reddie: But this was... He was from Aberdeen, Scotland. And people used to say, he's like an unofficial person who could be the successor to Deming. So he wrote a lot of wonderful papers. So David Kerridge, Henry, and I, were sitting together having dinner, and they were talking about Deming, and they said that, yeah, he did not... There were some things where he could get really, really upset about when he... About performance appraisals. But there are some things when he spoke of Shewhart, Walter Shewhart, he spoke in a very different tone. And then came this topic, that when we told him that you said reduce variation, he said, "I never said that." They said, "But you said it. We have it... We wrote it down, both Henry and I." And he said, "I could have said that in a context, but I don't think I meant it in that way." So for the record, yes, he uttered it; but he meant it differently. I'm quite sure he said to understand and manage that variation. 0:05:35.4 Balaji Reddie: So what is that managing variation? Okay. We saw this last time, in the very narrow sense of the term, that is when Walter Shewhart invented the control chart. He said there are two kinds of variation that exist: natural variation and unnatural variation. And to distinguish between these two, he created the control chart, which was based on... It was an empirical formula, and people confused it to be a statistical thing. He demarcated, brought in two lines, what he called as the upper control limit and lower control limit. And he said as long as the outcomes of any process or system, if you could measure them and you're plotting them on a graph, as long as they lie within these limits, well, that's the natural behavior of the process or system under study. And the moment their performance crosses the limits, one of these limits on the higher side or the lower side, then you can say there's something unique, special, that has happened. So he used the term, random variation, and assignable cause variation. Because he said you could assign a particular cause to the points that lie outside limits. But when you are studying the variation that lie within limits, the variation that lies within limits, then you can't ascribe a particular cause. There is a system of causes acting. This particular difference in approach... 0:06:58.1 Andrew Stotz: That's a great point. The assignable thing is a great, great point. It really helped me just understand it better. 0:07:04.7 Balaji Reddie: Oh, yeah. Because you could assign a particular, that's why he said assignable. And this is random. Assignable, has its advantage that you can isolate it. And initially, he said you could isolate, and you could eliminate, that cause because it is something which is alien to the system under study. And one by one by one, when he said you eliminate all those causes, you land up with a system or a process that fluctuates only within limits, and he called that as a statistically controlled process. Now, he said your real work begins when you start looking into the reasons for each of those points lying within limits, and I made that statement. You don't have one cause, you have a system of causes. Here's where we studied systems the last time, and I'm just refreshing all of this. I said that in a system, you have interconnectedness, you have interdependencies. And then cause and effect are not closely related in time or space. You have synergistic relationships. So when you start studying, it becomes very, very complicated. Because the performance of each of those causes is also subject to variation. So you can never, ever tell with absolute certainty, what is causing the system variation. Because each of the causes is part of a bigger system. So he made that statement that the complete knowledge of any process or system, is unknown and unknowable. He was not being pessimistic; he was being realistic. We discussed this the last time. And so I said, what do we do then? Yeah, we'll come to that. But now comes the other part, which gets really, really interesting. 0:08:46.7 Andrew Stotz: By the way, I want to... Before you move on from the variation, I want to add something. But continue on. 0:08:55.2 Balaji Reddie: Okay. So he said the reason for the point lying outside... And I said it's assignable, what Dr. Deming called as, special cause variation. You can assign a cause, and I said it's alien to the system under study. And I started by saying you can eliminate it. And even Walter Shewhart began by saying, eliminate it. But now, I'm going back on my words slightly. If you eliminate that special cause, it could be a special cause for the system under study, but it could be the common cause of another system which is related to this. And so what can you do? Well, here's the best part that comes in. You reduce the effect of that cause, to the extent that it does not cause any harmful variation here, to the system under study. And that, Andrew, is optimization. "I want this, and I want that", not, "I want this, or I want that." So I need that. 0:10:01.0 Balaji Reddie: Because I can begin with eliminating, but then I have to keep my eyes and ears op