Home > The Lean Post> The Secrets of Lean
The Lean Post
Sharing how the world is making things better through lean.

The Secrets of Lean

by Dr. John R. Ehrenfeld & Tom Ehrenfeld
March 24, 2015

The Secrets of Lean

by Dr. John R. Ehrenfeld & Tom Ehrenfeld
March 24, 2015 | Comments (10)

Lean manufacturing owes its success to its effectiveness in practice. This system of learning has come from experience, not theory. It emerged as a series of countermeasures designed to tackle immediate problems, and over time became codified, and recognized (this history is shared in the wonderful book The Birth of Lean) as a system. In this sense, it is a radical alternative to dominant “rational” management frameworks. The radical nature springs from a tacit assumption about the systems being addressed; they are deemed to be complex whether explicit or not. This is the key to their appeal, because virtually every system involving human actors is, per se, complex.

Complex stands over and against the dominant, modern worldview of systems as deterministic. Machine-like system, in this sense, can be described analytically and, consequently, can be managed by applying the rules. Some machines are very complicated and not completely known entities so that the rules leave some uncertainty in place. Complex systems are fundamentally or paradigmatically different. They cannot be described by analytic rules. Their future behaviors cannot be described within some degree of certainty. Unknown, unpredictable events occur at unforeseen moments. They can jump from one seemingly stable regime into another as the US financial system did in 2007-8. They cannot be known in the same way as machines.

But they can be understood such that agents can interact with them in order to keep them running within some distance from their optimum operating point. The key to Lean is that it creates understanding, not knowledge. One difficulty in successful implementation of lean systems is that managers conflate these two important concepts. Understanding and knowing exist within two distinctive philosophic domains, pragmatism and positivism, respectively. The methods to be applied in either one do not work in the other. Positivism only works in systems that can be described analytically, that is, as some form of formula or logical expression. Problems that arise are invariably handed to experts of one sort or another, that is, to people who presumably hold the knowledge required to solve the problems.

Problems in complex systems demand a radically different framework. Whereas positivism requires one to develop theories (knowledge) about the system at hand by standing outside of it and observing objectively, pragmatism generates understanding by living within the systems. By careful observation over time, with perhaps, a little measured perturbation now and then, observers can begin to develop a holistic sense of what causes the kind of deviations from the desired state, and thus, have non-random, ways to deal with those deviations when they inevitably occur.

The key players are those with the closest most intimate experience with the particular system of interest. They are polar opposites to experts who arrive on the scene with acontextual, abstract knowledge that only, by chance, fits the unique concrete nature of each and every complex system. No external hierarchy determines who should comprise such a group of observers/actors. Familiarity and concern about the system are the only criteria.

John Dewey, the great American philosopher of pragmatism, saw this to be a key feature of a sound political system that could deliver the values foremost in our Nation. The complexity of the society would be better governed by democratically derived solutions than by expertise. The involvement of those within the teams of inquirers/solvers is additionally empowering as the quality circles in the TPS has shown.

The last “secret” of Lean is that the most important outcomes from complex systems are frequently qualities like resilience; in the case of TPS, it is quality, itself. They emerge only when the whole system is functioning within some limited operating range. Merely tinkering with the system, that is, by applying analytically-based remedies cannot produce them. These qualities over emerge over time through understanding and careful interventions. Quality is not some engineer’s metric. Once again, TPS carefully involved its customers in its quality-producing program. Circling back, the key to success of Lean is the recognition of its radicality. It is not just some very clever method; it lives in a different belief system - a different paradigm - and managers that cannot understand this may have short-term success, but, sooner or later the old, entrenched beliefs will start showing up as new persistent problems. 

The views expressed in this post do not necessarily represent the views or policies of The Lean Enterprise Institute.
Keywords:  history,  systems thinking
Was this post... Click all that apply
13 people say YES
30 people say YES
23 people say YES
14 people say YES
Related Posts
10 Comments | Post a Comment
Jeff Liker March 24, 2015
3 People AGREE with this comment

What originally got me interested in the Toyota Production System was systems thinking.  At that time it was coming out of sociotechnical systems.  This is a great essay on systems thinking.  The distinction between knowledge and understanding is profound.

Reply »

John Hunter March 24, 2015
4 People AGREE with this comment

I think you make good points, but I think you make a mistake stating

"This system of learning has come from experience, not theory.'

For some reason, culturally, we created this idea that theory was about disengaged people (away from the gemba) thinking in a way disconnected from practice.  But this is not what theory is.

Learning can't take place without theory.  Experience doesn't lead to learning.  Experience with a theory and an informed observer that questions what they see can lead to learning.



Reply »

Dirk van Putten March 25, 2015

Hello John: I almost posted essentiallythe same comment you made about learning requires theory.

Reply »

Jay Bitsack March 25, 2015
1 Person AGREES with this reply

Hi All,

I concur with John...  Lean thinking and behaving as a complex "adaptive" system manifest in the form of TPS/Toyota Way has a very special property/characteristic that is highly "knowledge" dependent.  More specifically, in order for the system to continually operate as intended at any specific moment in time, it's ability to do so and remain within a tightly balanced set of operating parameters is highly dependent upon the existence of a body of common/share knowledge which is readily accessible and leveragable by all those individuals who are acting as elements within the system.

In contrast, because the system is adaptive not only to changes in conditions that occur/arise in the present, but it is also adaptive in an anticipatory sense.  And that ability to adapt itself and emerge/evolve with a new and improved set of internal competencies and capabilities is NOT predicated on EXISTING knowledge, but rather on NEW knowledge that has yet to be generated and created.  And that new knowlege is needed/intended to expand on existing knowledge and does so via the application of theory, followed by experimentation - which is used to disprove or substantiate and/or refine the theory as necessary into a newly evolved working body of knowledge.


For anyone who is interested in the notion of complex adaptive systems (CAS) and how they operate, one needs to go back to their origins in 1984 at what called the Sante Fe Institute - founded by a group of distinguished scientists including Nobel Prize winner Murray Gell-Mann.  Eight of the ten founders were scientists with the Los Alamos National Laboratory.  The initial intention of the members of this organization was to overcome the "reductionist" thinking that had (and still does) permeated science and academia and focus on discovering "the new rules" of underlying order of the universe (and the sub-systems that exist within it) that chaos theory implied.

However, by 1990s, after finding that the study of chaos was too chaotic an undertaking, the members of this organization turned their attention to studying complexity.  What they came up with was the notion and science of complex adaptive system (in contrast to complex non-adaptive systems such as the weather on planet earth).  And by definition, complex adaptive systems have the unique ability to internalize information (obtained from all different types of sources - both internal and external) in order to learn, and subsequently modify their behavior (or evolve) as they adapt to changes in both the existing  and anticipated future environments.


Bottom line: Knowledge and understanding are intimately connected.  In order for any human being to understand anything about something, there must first be a body of existing knowledge to work with.  And how that existing body of knowledge expands/evolves is greatly dependent upon the degree/level of understanding [regarding that body of knowledge and other potentially related bodies of knowledge] that an individual (or group of individuals) possess.  At this point, we'd have to delve into the area of D.S.R.P. thinking to make a deeper connection.  But suffice it to say that more information about this broader form of thinking ability (which incompasses systems thinking) can be obtained by Googliing it or Dr. Derek Cabrera.

Reply »

Jay Bitsack March 25, 2015
1 Person AGREES with this comment

Hi All,

Forgot to mention... One of DEEPEST and least understood/appreciated secret of true lean thinking and behaving is the fact that much of its on-going successful evolution/emergence as a system is predicated on learning from FAILURE; that is by pushing the system and/or its components/elements beyond their theoretical design limits to the point of failure.

OMG - that sounds like lean thinking heresy!  Right?  NO... ABSOLUTELY WRONG!  Toyota Motor Company - from its founding days - has held fast to the belief in and practice of pursuing/creating new knowledge.  It's a legacy that was given to the organization by two of its earlest founders... Sakichi and Kiichiro Toyoda.  They both recognized from the get go that in order to make progress and remain ahead of the pack in a highly competitive environment, that one must never become comfortable/complacent with what one knows and understands.  Instead, one must continually search for new knowledge and learn how best to leverage it based on one's deepest understandind of it.

In this regard, when it comes to operating/proving-out any new system/process intervention or product componet to the point where it operates as desired or needed within the system it is a part of (i.e., compliance with specification) only proves/demonstrates that the existing knowledge used to create it was correct/valid within the boundaries of the targeted performance.  What it does NOT do is provide any indication of what new knowledge may be needed in order to make/pursue further improvement(s).  And the only way to obtain such new knoweldge is to push beyond/outside the boundaries of existing knowledge; ergo the need for and practice of pushing/testing/experimenting to the limit of falure.

Reply »

Rick Bohan March 26, 2015

One could argue that much, maybe most,  learning (behavior change) is from failure (that didn't work...I'll do something else next time).

Reply »

Rick Bohan March 26, 2015
4 People AGREE with this comment

I don't mean to be unkind but a less readable, more abstruse essay is tough to imagine.  What do sentences like the following even mean?

"Complex stands over and against the dominant, modern worldview of systems as deterministic. Machine-like system, in this sense, can be described analytically and, consequently, can be managed by applying the rules."


“But they can be understood such that agents can interact with them in order to keep them running within some distance from their optimum operating point.”

Or, best of all…

“By careful observation over time, with perhaps, a little measured perturbation now and then, observers can begin to develop a holistic sense of what causes the kind of deviations from the desired state, and thus, have non-random, ways to deal with those deviations when they inevitably occur.”

And all this is apart from the fact that the authors’ premise is wrong from the start.  They set up a straw man argument that nobody ever seriously makes, that one can come to understand a system without ever really immersing oneself in it, and then rebut it with lots of rambling prose straight from a thesaurus.

Knowledge and understanding aren’t different things.  One can’t reasonably say, “I have a great deal of knowledge about that process but I don’t really understand it.”  Nor can one say, “I fully understand that process even though I don’t have much knowledge about it.”

They do manage to land on the truth with the already widely accepted axiom:

“The key players are those with the closest most intimate experience with the particular system of interest.”

The authors won’t have any problem getting readers to agree that those closest to a system or process generally have a good deal of understanding and knowledge about that process.  But, geez, guys…couldn’t you have said that right up front much more clearly than you did?

Reply »

Jeff Morrow April 01, 2015
1 Person AGREES with this reply

It's chewy, that's for sure.

But I have sympathy for the author, too, because he's had to shoe-horn some (elsewhere) very well elaborated ideas into a short-format piece. Pretty much a given that when you try to do that you're gonna have highly specific, atypical expressions.

Further, I think it was worth the effort because the contrast between the worlds of positivism and pragmatism and is profound but only subtly separates Sloanite-Welchian industrial organization from Toyota Way.

Another example of this more or less visible schism can be seen in North American statisticians' response to Taguchi's Robust Design. The statisticians, Positivists par excellence (thank you, Sir Ronald!), viewed Taguchi's pragmatics with alarm and disdain - they just couldn't grok the idea that you don't need to rigorously characterize parameters' effects and their interactions to take advantage of thoughtful experiments exposed to the complex wildness of the Gemba.

Reply »

Tom Ehrenfeld April 01, 2015
1 Person AGREES with this comment

Thanks for your comments!

We believe that the apparent differences in the arguments about theory are largely semantic and not substantive. We were using theory in the positivistic sense, as most everyone does. Theories, in this sense, are a priori. hypothetical analytic relationships that are established as valid through controlled experimentation. Understanding is developed by a different process in which beliefs about a complex system are created a posteriori, that is, after the fact of observing the system. Given the innate, indeterminate character of complex systems, controlled experiments are virtually impossible as the boundaries of the system are always poorly defined. Learning certainly does take place; that is the point of pragmatically based inquiry frameworks, but it is of a different kind than theoretical, positive knowledge. If the experimental methods have been carefully applied, the resultant knowledge can be expected to lead to predicted results when applied to a essential similar situation. Not so with understanding; it is always contextual and has meaning only for the system from which it was derived.

Systemic understanding, created by pragmatic inquiry processes, such as those found in the TPS, is shared by the inquirers. What is understood certainly is used to guide corrective action to dissolve the problems at hand, but this is not some theory that can be generalized and applied willy-nilly. The understanding may be found to useful if the same circumstances arise, but, if the learning was correct, the particular problem should vanish. The results of a pragmatic inquiry cannot, with perhaps few exceptions, be reduced to an analytic expression, one of the defining features of conventional theories.

Individuals learn from experience by restructuring our brains. If one argues that action comes from some sort of "theory" embedded in the brain, then the commenters' statement would be correct, but that is not our understanding of the way most neuroscientists talk. We interpret our perceptions (experience) through our pre-existing beliefs or "prejudices," as H-G. Gadamer writes. In a sense, our beliefs are cognitive structures with which we make sense of the world and which subsequently guide our actions. Yes, they could be called theories, but without much qualification, this usage would not be helpful. I don't think it would work to say, "I have a theory that the sky is blue." Most people would simply raise their eyebrows. No such response comes from my saying, “I believe the sky is blue.” I think the comments and our response indicate that we are not so far apart. The chief reason for making the distinction, we believe, is that when the word “theory” or “knowledge” comes up, it immediately places people in the context of analytically describable systems, when they should be thinking about complex systems that demand a distinctly different frame of reference.


Tom and John

Reply »

Jay Johnson May 01, 2015
1 Person AGREES with this comment

Interesting topic for sure. Are you familar with the work of Dave Snowden with Cognitive Edge?

Reply »

Lean Lessons from Tesla
Lean Thinking: A Roundup
Do CEOs Matter?
Happy Birthday Eiji Toyoda!
Please include links as plain text URLs only. Do not copy and paste directly from a web page or other document. Doing so may pick up additional HTML that will not function here.
URLs will be converted to functioning links when your comment is displayed on the site.
Here's an example:
See this article for more details: https://www.lean.org/whatslean