Correcting Goldratt's Cloud - We Are Doing It All Wrong: Part One

Introduction

Goldratt's cloud should teach and liberate; be systemic, local/global, and tease out the paradox that will take us forward. Goldratt's cloud should not torture and bind; be mechanistic, local/local, and immerse us in a dilemma from which we cannot escape.

I haven't drawn clouds since November 2014 and that is simply because I found that a systemic matrix is easier to follow and parse out both the current problem and the direction of the solution; after all, it's all there in one place. Moreover, it has something that the cloud hasn't. It has something that the cloud hasn't but should. In writing the previous article on Theory of Constraints is Obsessed with Dilemma and this is Wrong, I realised that clouds are missing something. I missed it too. The cloud doesn't have the explicit UDE's mapped into it. We need to correct for this. I'll show you how. Moreover, by correcting for the presence of the UDE's we will also discover why dilemma clouds oscillate and clouds of paradox stagnate (at first). But we should really correct the whole cloud while we are at it. We need to learn how to do paradox, not dilemma.

The following is in two parts. This, the first part, is really quite generic, I'll often use generic terminology which requires that you have some prior knowledge of Goldratt's Thinking Process. But I will also illustrate it in the second part with the big batch/small batch - um - issue, so that there is something concrete to hang things off if you aren't so familiar with this process and terminology.

In this first part I will show; clouds for dilemma and for paradox, some recent history of how the matrix and the cloud are related, and then develop things as far as what is missing from clouds in a generic sense. We are going to put the UDE's back in.

In the second part, I'll delve back into some deeper history - to when I didn't know what I was doing was different - and examine the specific case of big batch/small batch and "work" it - as it should - as a systemic cloud and show the paradox. That will help expose the underlying blueprint or template for any systemic or paradoxical solution.

Background

If we go back to Goldratt's earliest clouds, we will find something like this.

Excuse all the white space, it will get filled in later; I just want to keep a consistent scale to the diagrams. This is from page 43 of Goldratt's (1990) What is the thing called Theory of Constraints: and how should it be implemented. I've reversed the upper and lower arm to maintain consistency - you will see why. A page or two prior, page 39, are two generic examples. So, let draw a rendition of the generic issue here.

This is a dilemma cloud, a local/local cloud. The two "wants," the D and the D' (D prime) are at the same logical level - check back to the example above, it's about batch size. AND the two "needs," the B and the C are also at the same logical level - this time it is about cost. We can read these back as:

In order to have B local need we must have D local want

and

In order to have C local need we must have D' local want

the key phrase is:

In order to ... we must ...

The B/C of the cloud, the needs, the requirements, are always at the same logical level; two alternatives. Naturally you bounce around between the two; you oscillate from one to the other and back again.

Now, I didn't pick that earliest cloud as a trap, or a straw man, that Goldratt later evolved away from. I picked it because it turns out to be formative for everything else that follows after it. It's like we went down one evolutionary branch and there was no coming back. We see that same family if you like, well even the exact same clouds, reproduced in the (1996) Production the TOC Way simulator manual, and again in the 2003 revised edition.

So, what is the other path? Well, it looks like this.

Now, have I just trapped myself? You see, we have been down this pathway or so it seems as there are several clouds in the production manual that have this general form. In fact, let's draw one.

That's from page 84 of the instructor's section of the manual. Once again, I've swapped the upper and lower arms. Now, both of these; this one and the one I drew before, have the same form. Let's see.

This is a systemic cloud, a cloud of paradox, a local/global cloud. The two "wants," the D and the D' (D prime) are still at the same logical level - check back to the examples above, they are still about batch size. HOWEVER, the two "needs," the B and the C are now at different logical levels. The upper one, the higher logical level, is about lead-time or utility (to the customer) and the lower one, the lower logical level, is about production or productivity, which is a singular and internal issue. We can read these back exactly as before:

In order to ... we must ...

Here, the B/C of the cloud, the needs, the requirements, are not at the same logical level; one is higher than the other. Naturally you stagnate at the lower logical level. You need to recognise this. This is a paradox. You want to go forward, but you stay still at the same time. You want to diet but that chocolate cake just got you once again (dang - and the time before that as well!)

Now the trap (that isn't). I argue that Theory of Constraints is obsessed with dilemma, and that dilemma takes us down the rabbit hole, but it is only paradox that gets us out again. And yet I've just shown you a Goldratt cloud from the production manual that is apparently systemic. Well, it is, systemic - and not only is that in italics, but you should also see how I pounded that out on the keyboard. BUT here's the thing. Nobody recognises this!

That is not to say that there are not systemic clouds within Theory of Constraints. Let's take the double negative out of that and start again. There are systemic clouds within Theory of Constraints, but they are not recognised as such.

There are a number of B/C pairs in Theory of Constraint clouds that are paradox or systemic clouds, but they are treated as dilemma. The one that always springs to mind is Efrat's cloud of security and satisfaction. It isn't security vs. satisfaction, it is security and satisfaction. We don' give up security to get satisfaction, in fact, we enhance our security by working cooperatively at a global rather than local level - but that's another story. Security is the lower logical level and satisfaction is the higher logical level (as the problem is presented).

In the previous article or the comments that developed out of that I explained a test. Dilemma clouds, local/local clouds oscillate. Go back to the first one at the top here on costs and you will see how that might be. Clouds of paradox, local/global clouds stagnate. Check the clouds directly above. We love to make big batches. The bigger the better. We hate to make small batches - and that remains Toyota's competitive advantage to this day (even though small batches are better - which is the paradoxical solution we will come to).

Think about it a bit more. Theory of Constraints is founded upon the idea of constrained production. Nothing about utility, little about lead-time, nothing about quality. Give us something to (instinctfully) exploit and we are all gung-ho, give us something to (thoughtfully) subordinate and we are a quivering mess. We stagnate on the lower logical level of exploitation and fail on the higher logical level of subordination.

New York 2011

At New York in 2011, I did my first presentation at a Theory of Constraints International Certification Organization conference. I presented on systemic clouds - something that I had been doing automatically since 1998 without knowing that I shouldn't. In fact, I had the temerity to think that others should - and clearly, I still do. I had naturally been drawing clouds as systemic clouds ever since I first learnt how to draw clouds. I tried to encapsulate the key aspects of those in a 2008 PowerPoint on my website. I'll redraw them in the second part of this discussion.

Around 2014 I became quite conscious that I had not been drawing local/local clouds at all - or as a North American colleague called them loco loco clouds - nor had I had any need to; we are dealing with people in systems and our problems are systemic. So, I did draw a did a local/local cloud at that time - its Neil Diamond's I Am I Said. You can find those lyrics and the original music set to a cloud on the opening page of my website www.dbrmfg.co.nz. That song has always intrigued me. I knew there was a cloud in there somewhere. I also did that as an exercise, in part, simply to show to myself that a set of rules that I had developed in 2012 for systemic clouds applied to local clouds too. Let's have a look at those rules.

Chicago 2012

At the 2012 inaugural (and only) Goldratt Foundation New Knowledge Awards, held in Chicago as part of the annual Theory of Constraints International Certification Organization conference, I presented on the relationship between Goldratt's cloud and the change matrix, showing the relationship between the two.

I knew the matrix was systemic since its YouTube debut in July 2010 as a teaser for the then upcoming book Isn't it Obvious. As I've related elsewhere, I immediately mapped Goldratt's scheme for resistance to change onto this. This is how it looked.

Now remember that Goldratt presented this as what I would call a "flatland" version." If we go back to Goldratt's 1996 My Saga (Ah! I found my printed version - its pp 1-14 in Production the TOC Way - I knew that I had it somewhere!) then this is what the scheme first looked like.

The layers of resistance are:

1. The first layer of resistance: raising problems having one thing in common - it is out of our hands: vendors do not always deliver, clients change their mind at the last minute, workers are not properly trained, corporate forces on us ...
2. The second layer of resistance: arguing that the proposed solution cannot possibly yield the desired outcome.
3. The third layer of resistance: "Yes, but ..." Arguing that the proposed solution will lead to negative effects.
4. The fourth layer of resistance: raising obstacles that will prevent the implementation.
5. The fifth layer of resistance: raising doubts about the collaboration of others (or worse, not raising their doubts).

I tend to short-hand that to something closer to this:

1. Problems
2. Solution
3. Reservations
4. Obstacles
5. Unverbalised fear

Now, I said that was a flatland version, because I believe that, at least originally, Goldratt was absolutely aware that this was an adaptive challenge. He said as much in his 1990 version in What is this thing called Theory of Constraints (pg 10)

• Any improvement is a change.
• Any change is a perceived threat to security.
• Any threat to security gives rise to emotional resistance.
• Emotional resistance can only be overcome by a stronger emotion.

That cries out adaptive challenge. If you overlay that on the 5 layers of resistance to change, the step, the challenge, lies between the problem and the solution. That why I drew it as I did above on the day I first saw the change matrix. And again, as I've related elsewhere (Of Mermaids and Alligators), I had some exposure to these step-like changes and matrices some 13 year prior with the work of Aubrey Daniels.

To my mind, Goldratt tried to "walk back" the layers of resistance, renaming them buy-in and later de-emphasising resistance as much as possible - "would you resist winning the lottery?" was one of his ploys. Whereas Theory of Constraints rejected resistance, Robert Kegan embraced it as immunity. That rejection, in fact, is consistent with a view of change as a technical challenge, or as a dilemma.

Okay, lets rule a line under that.

Back to the discussion. Firstly the "how" could I relate the cloud and the matrix? And the answer is that because the clouds I draw are systemic, and always have been, and so too is the matrix. Now, to the "why" did I relate them?" I came away from the 2011 conference (Goldratt had passed away a short time prior) with the understanding that he thought the cloud and the matrix were related but he didn't know how. I would take that as another line of evidence that Theory of Constraints was immersed in dilemma or technical challenge by that time and not paradox or adaptive challenge. That's two reasons - I'll summarise them in a moment. There is one more.

There was a spirited opposition to what I am about to show you that came from a number of members of the committee that administered that prize. To the extent of publicly attacking the presentation at is completion when they had been in the privilege position of knowing the content for many months and had entertained a secondary rebuttal of some of their earlier objections. That's three, lets summarise this.

It would seem like paradox and adaptive challenge were not part of Theory of Constraints and that paradigmatically, and to the exclusion of all else, dilemma and technical challenge was part of Theory of Constraints.

• Goldratt had rejected resistance to change.
• Goldratt couldn't relate the change matrix to his dilemma cloud.
• Goldratt Foundation Committee opposed a systemic connection.

Of course, there is also the fact that clouds were, and still are, exclusively interpreted as dilemma.

Let's have a look at what the fuss was all about.

The Cloud and the Matrix

Here is a generic version of the cloud that I presented in 2012.

It looks, well, rather pedestrian these days. The key point was the:

"because of the positives of this side,"

"and because of the negative of the other side"

because they came from my interpretation of the matrix at that time. This is what that matrix looked like in essence.

Look carefully. Each of the diagonal in the matrix has a positive of one side and the negative of the opposite side - and that diagonal pair occurs together on one side of the cloud above. Go on, trace it with your finger. Check it.

Now again, that kind of looks pedestrian today - but at the time - August 2011 when I first did it, it took a concentrated effort; to relate the cloud and the matrix, to confirm Goldratt's feeling that they were related even if he didn't know how.

But now things get more interesting. You see, not only can we put the matrix in the cloud or relate the matrix to the cloud as we have just done, we can also go the other way. We can put the cloud in the matrix (There is a Cloud Hiding in the Matrix - and it Isn't Obvious). This is how it looks.

Now we are getting somewhere. Check this diagram with the generic clouds further up. Both have needs, and both have wants. Both have DE's (desirable effects), BUT only one - the matrix - has UDE's (undesirable effects). The cloud only has a positive expression of that, which is to "avoid the UDE's of the other side." I realise as I do this that I am using DE/UDE and positive/negative interchangeably. Just go with the flow. But we've got to an important point.

Where Are the UDE's in the Cloud?

So, where are the UDE's in the cloud then. It's a kind of a where's Wally jigsaw for grown-ups. Let's have a look.

Okay, I have DE's or assumptions about DE's and I have the avoidance of UDE's of assumptions about them in place, but still, where are the UDEs in actuality? I know where we are expressing them - on the other side from where they are. But where in the heck are they? And when I go to that side it actually says go back to the first side. Let me show you.

And

So,

Well, the late Prof. Antoine van Gelder knew a thing or two. He shared many a late-night discussion with Eli Goldratt when Goldratt was working in South Africa. Antoine also put the jeopardy arrows into the cloud. Let's take the conflict arrows out, and put the jeopardy arrows in.

Now, let's stop there for a moment. If I go and get the 2010 Theory of Constraints Handbook, that 1175-page doorstop, and I count through the number of clouds drawn in there, I get around about 110 - give or take a few. And not a single one has a jeopardy arrow in it, let alone a pair. In fact, there is no mention of jeopardy arrows at all in that tome.

****

[I wondered if I could find any evidence of when I first knew of jeopardy arrows. I found that in a small collection of CMSIG (APICS Constraint Management Special Interest Group) posts that I had "saved" in Excel from August-September 1998. There is a single jeopardy arrow (B-D'), rather than a pair, in a series of follow-up in posts by Tony Rizzo, Larry Leach, and Rob Newbold to a post by Ted Hutchin, "Hutchin's Paradigm Lock Cloud." Why did I go to the effort of drawing these out and typing them up in Excel? Well squirrels do that. But these 4-5 posts were about clouds which certainly could be viewed as having different logical levels and I guess that is what resonated with me.

And let's push the point, Rob Newbold and Larry Leach both wrote acclaimed books on Critical Chain, and Tony Rizzo had a seminal influence on multi-project Critical Chain that most people use today; and here they are thoughtfully and freely discussing systemic clouds, or local/global clouds, or clouds of paradox, but the instinctful orthodoxy of Theory of Constraints has never embraced that wisdom.

We deal with systems and yet the orthodoxy wants to frig around with dilemma!]

****

Okay, here is some redemption or sorts. Jelena Fedurko in her 2011 book Behind the Cloud: enhancing logical thinking - which let's be very clear is about dilemma - has a chapter on logical cross-check: "D endangers C", "D' endangers B." Those are indeed the jeopardy arrows, there being used to check the rigor of the needs (B/C) and wants (D/D') of the cloud.

But back to jeopardy arrows themselves. Here's the thing. The UDE's are the assumptions under the jeopardy arrows. Let's draw them in.

Does that make sense? I hope it does. That's where the UDE's are.

I can understand why I originally drew these as "and because of the UDE's of the other side," it was out of respect for Goldratt's intent that the cloud should be profoundly positive - and out of that positivity comes these perplexing negatives. Avoiding a negative or an undesirable effect (a double negative in effect) becomes positive and preserves the overall intent of the cloud.

These assumptions/UDE's are really, really, important. They are the anomaly that we drew in the lower-right and upper-right of the matrix. I'd drag that diagram back, but you will have to page-up instead because I don't want to interrupt the flow. No, I will, here's the diagram again.

Once again, those UDE's are so important. They are the anomaly that tells us something is wrong. The are the reason we go looking for the something that is right - the enabling assumptions at the new and higher level.

Now, if we put the UDE assumptions in the cloud, then aren't we violating Goldratt's positive intent? Well, the answer is a resounding no! In fact, these are the indicators of the direction of the solution. Without them we cannot climb out of the hole. Let me show you.

Historically, we read jeopardy arrows - or at least I do - as:

D jeopardises C

and

D' jeopardises B.

Banal - right? Clearly it begs expansion as follows:

D jeopardises C because of the underlying UDE's,

and

D' jeopardises B because of the underlying UDE's.

But now we have a negative - a jeopardy - in Goldratt's positive cloud. How do we get out of that?

Well, by returning to the original language of necessity logic. Watch this.

In order to have C we must not D because of the underlying UDE's,

and

In order to have B we must not D' because of the underlying UDE's,

Do you see, must not creates a double negative and restores the positive intent. The B-D and C-D' arrows or assumptions or DE's are things we must do, and the B-D' and C-D arrows or assumptions or UDE's are things we must not do.

Summary Part One

I hope that I have shown you three important things

1. There are jeopardy arrows in the cloud.
2. There are UDE's in the cloud as well and they form the assumptions that are "under" the jeopardy arrows
3. It is the jeopardy arrow that drive the dynamics of oscillation in dilemma and stagnation in paradox.

Like so much in TOC - all of that "stuff" was hiding in plain sight for all to not see.

Now, there has to be a paradoxical solution - otherwise that cloud holds, and we stagnate in the lower arm.

Let's save that for part two.

A Word About the Banner

Searching for artwork on paradox or paradigm or logical level for the banner, I settled for something quite personal and indigenous instead. "The koru (Māori for 'loop or coil') is a spiral shape based on the appearance of a new unfurling silver fern frond. It is an integral symbol in Māori art, carving and tattooing, where it symbolises new life, growth, strength and peace." That's what Goldratt's cloud ought to be doing too.