Part Two: The Deep Irony of Taylor, Ohno, and Goldratt

Introduction

Once again, let's cut to the chase; in operations the real productivity and prosperity is in the non-constraints - the real productivity and prosperity is in the non-constraints. How could it be any other way? They might be the only thing that you have. It is exactly the same situation as with quality where the real signal is in the noise (see Part 1: The Deep Irony of Walter Shewhart and W. Edwards Deming). I can assure you that there is both a deep irony here and more than a little apparent contradiction.

In quality we saw the virtue of restricting our tampering from everything everywhere all of the time - which only made matters worse - to addressing only those few significant instances as identified by the natural process limits. Essentially, we applied Shewhart and Deming's thoughtfulness to rein in, or restrict, our own instinctfulness. The vice and indeed the irony is that we then failed to apply any further thinking, any further thoughtfulness, especially our own, to the untampered variation that remains. In fact, I drew this view of what we do (virtue) and what we don't do (vice).

When we do apply our own thoughtfulness to the remaining untampered variation, or we follow Deming's lead, we find that there are just 4-5 factors that account for the vast majority of the observable effects. This happens at two levels: dominant general causes and lesser causes. We feel as though there are a plethora of factors to account for, but in effect, there are only a relative few and they account for, by far, most of the observed variation. We can therefore make good and rapid progress.

Let's now transfer this understanding from quality to operations and logistics. I mentioned in Part 1 that we can boiler plate from one to the other, but it is more that too. I developed each of these ideas - about quality, and then about operations/logistics - quite independent of one another until it was inescapable to me that they were following the same rules, the same meta-methodology. For the record, it was operations first and quality second. In fact, it's the same interaction between our instinctfulness and our thoughtfulness in both cases. It's like going cross-eyed and seeing two, then snapping back into focus and only seeing one (the meta-method). The ability to boilerplate from one to the other and back again is simply further evidence of the underlying/overlying commonality between each of these and the meta-method.

Let's start our investigation of operations/logistics in a much early era with Frederick Taylor and what he eventually came to call Scientific Management.

Understanding Tayor's Understanding is Quintessential

Margaret Wheatly wrote with reference to Taylor, Gilbreth, and their host of followers that: "I still find this early literature frightening to read. Designers were so focused on engineering efficient solutions that they completely discounted the human beings who were doing the work. They didn't just ignore them, ... they disdained them - their task was to design work that would not be disrupted by the expected stupidity of workers."

Except she missed the mark by 100 miles and almost as many years, although I will give her a point for Taylor's later hyperbole (and given the concepts and things that he invented - including high speed steel for instance - he might just be forgiven for that). She continues: "Though we in management may have left behind some of these beliefs and the rigid, fragmented structures that those beliefs engendered, we have not in any way abandoned science as the source of our credibility."

Well, at the risk of a huge divergence from my initial intent, when I first read that more than 25 years ago, I was aghast. To quote Deming: "It is not enough to do your best, you have to know what to do, then do your best." These guys - the workers - weren't stupid - and nor did anyone else think so, they were simply doing their best in the absence of knowing anything better. And hell's bells, their "management" knew no better either. Science wasn't the issue; it was the absence of science that was the issue. An absence of science and an absence of thoughtfulness that Taylor sought to rectify.

He sought to make work easier, not harder. I'll repeat that because people want to disbelieve it: he sought to make work easier not harder. The Gilbreths likewise. You can read about Henry Noll or Knoll, or Knolle, and who by the time of Scientific Management had morphed into "Schmidt," in either Kanigel's biography of Taylor (The One Best Way: Frederick Winslow Taylor and the Enigma of Efficiency) or Taylor's own 1911 work The Principles of Scientific Management. But here is the salient point; Taylor tried not once, not twice, but three times to find a relationship - a mathematical relationship - between manpower expended and work done. And on the third attempt he and his mathematician assistant found a relationship - it was rest. How ironic is that? The harder the task, the more a man was under load - in this case carrying pig-iron - the longer the rest in-between each exertion. Taylor ensured that relationship by adjusting the time taken to walk back unloaded before the next load. At the end of the experiments Henry Noll could carry 47 tons of pig iron a day by stopping work. That's not Margaret Wheatley's version, but that is what actually happened. And BTW the more a man could do in this time of piecework, the more he earned.

We are going to see time and again that rather than being locally efficient, being locally effective, that is stopping, is globally far more effective, and far more productive overall.

Taylor and Taylorism, more often known today as Scientific Management, was a monumental step forward and widely taken up in Europe and Japan in the following years. Let's now skip a couple of decades and have a look at the next significant change to operations/logistics.

Toyota

Let me run through a couple of comments first - because I delved back into Womack and Jones (Lean Thinking) and reminded myself that Jeffery Liker's (2004) The Toyota Way is an excellent Western interpretation. I understand that Mike Rother's 2009 Toyota Kata: managing people for improvement, adaptiveness and superior results is also very good, but I haven't read it. There is also one other source that I think is significant and underrated: Edwin Reingold's Toyota: people, ideas and the challenge of the new. It happens to come from a Penguin series titled Japanese Business: the human face. The human face - imagine that.

Also let me place some further context here. I worked in Japan for four and half years. The latter part of that was to put in place Theory of Constraints' drum-buffer-rope in a large and significant Toyota supplier. Imagine my amusement as I started to sing Taiichi Ohno's work, chapter and verse, in a Japanese firm, and a Toyota supplier at that.

But there are important things in the West that are often misunderstood about Toyota. Here are a few.

Andon. My own experience in Japan was with machine centres rather than flowlines - although each section of machines is indeed part of an overall flow - each machine had an andon light; green for operating, orange for setup, and red for stopped. I do have a photo of a large group of constraining operations all with green lights on. The factory had never experienced that before. But Ohno's intent was a little different.

"When a worker wants to adjust something on the line and calls for help, he turns on a yellow light. If a line stop is needed to rectify a problem, the red light is turned on. To thoroughly eliminate abnormalities, workers should not be afraid to stop the line." (Ohno pg. 121) My bold stress. BTW, didn't Deming say drive out fear?

It is quintessential (there's that word again) that for productivity to go up, the line must be stopped when there is a problem. I also know of this termed as an andon cord - a line stop cord.

Kanban. What does do the reverse flow of information cards do? Well, they pull work forward, but that was not their original intent. "The ... answer to the question "why are we making too many parts?" was "Because there is no way to hold down or prevent overproduction." This led to the idea of visual control which then led to the idea of kanban. (Ohno pg. 18) It's not a pull mechanism per se, it's a stopping mechanism. Can you see where this is heading?

Let's push on.

Takt/Tact. Continuing in the same vein; "To prevent overproduction and make items as needed, one by one, we have to know when they are needed. Thus, the appropriate tact time becomes important." (Ohno pg. 60). Do you see right there, that takt time has the job of preventing overproduction, stopping overproduction. Maybe not stopping in the sense of start-stop-start, but still, it is a stopping mechanism. Now, the reason for my dive into Womack and Jones. Here is their take on takt (pg. 56).

"Transparency facilitates consistently producing to takt time and alerts the whole team immediately to the need either for additional orders or to think of ways to remove waste if takt time needs to be reduced to accommodate an increase in orders." Oh really?

Do you see how that little snippet is totally at odds with Ohno's intent?

Jidoka. Just one more, a little more obscure, but Ohno raised the point (pg. 18) so let's run with it.

"Why can one person at Toyota Motor Company operate only one machine, while at Toyota textile plant one young women oversees 40 to 50 automatic looms?" By starting with this question, we obtain the answer "The machines at Toyota are not set up to stop when machining is completed." From this, automation with a human touch developed."

Do you see, once again, that word - Stop.

And Deming Too

It wasn't my intent to drag Deming into this discussion on operations and logistics, but he too knew only too well the value in a system of the obligation of a component.

"The obligation of any component is to contribute its best to the system, not to maximize its own production, profit, or sales, nor any other competitive measure.? Some components may operate at a loss to themselves in order to optimize the whole system, including the components that take a loss." (Deming, New Economics pg. 97)

He didn't quite say stop, but the intent is quite clear. And from reading John Willis 's Deming's Journey to Profound Knowledge, I'm guessing that there is a strong input from Ackoff that found its way into that part of Deming's 1994 edition.

Theory of Constraints

Before we investigate Theory of Constraints as the third entity in this evolution from Taylor to Ohno to today, let me say that the published literature on Theory of Constraints manufacturing solution drum-buffer-rope (DBR) is scant. Most of the DBR books are written by people who taught operations research (OR) at a university level, and we should be grateful for that. I certainly cut my teeth on Umble and Srikanth's wonderful 1996 volume Synchronous Manufacturing: principles for world-class excellence. I have owned and lent more than a few copies of that until I learnt that they don't come back. In the foreword of that book Goldratt notes the sequence of names for his method that proceeds from optimized production timetables in 1979, to optimised production technology in 1982, to synchronous manufacturing in 1984, to Theory of Constraints in 1987.

With respect to practitioners however, Mark Woeppel wrote what I consider to be the only practitioners' version of DBR in 2001: Manufacturers Guide to Implementing Theory of Constraints. I had the good fortune to meet Mark several times in Japan and several times since. And that book, by the way, is getting an update as we speak. This paucity of material is also what drove me to write a website on each of the logistical solutions and the Thinking Process after that period of working in Japan. You can find it here at www.dbrmfg.co.nz

In much the same way as with Taylor and with Ohno there are a couple of things that I believe are salient to Theory of Constraints but are largely ignored. I will develop the argument further, but for the moment let's look at two of these.

Material Release. Drum-buffer-rope (DBR) is essentially a pull system. It's pulls via time - the planned shipping date - rather than by a physical quantity as would kanban or a physical line as would a classical Ford system. One of the most important parts of this is material release. A material release schedule is simply the drum or shipping schedule offset in time by the duration or "length" of the buffer. But here's the thing; the material release schedule stops you from releasing material ahead of time. Or I should say, it ought to, but it won't until you have learnt your own lesson there. Been there, done that, only once, still got the scars.

It's a platitude, but putting material in sooner does not cause it to come out sooner, quite the opposite. Putting it in sooner only results in it coming out even later than you anticipated, and that just starts-off a horrible vicious cycle.

Roadrunner. This will be the one aspect of DBR that you will not have heard of. If you have heard of it, mention it in the comments. It gets significant attention in Goldratt's manual Production the TOC Way (1996 and 2003), it gets a paragraph in Schragenheim and Dettmer's 2001 Manufacturing at Warp Speed: optimising supply chain financial performance, and that is about the sum total. I give it attention on my website - Google "Roadrunner TOC" and you will find it. In fact, I was presented with a Tokyo Disneyland "Roadrunner" figure which used to sit on my desk.

Curiously, (to me) Goldratt does discuss this need in Haystack Syndrome - his 1990 write-up of his older OPT mechanism. He knew of the required subordination action; yet he clearly didn't at that time have a satisfactory name for it (in 1990). Let's see:

"Suppose we have already developed all the needed tools for the new work ethic. Companies that have already made the switch have found that, contrary to common belief, workers look for work. Some companies have even gone to the extent of putting newspaper stands on the floor; this didn't help. The idle time that is an unavoidable result of forcing non-constraint resources to refrain from overproducing must be filled with something meaningful." (Haystack pg. 89)

Firstly, I have to point out the Western paternal bias: "contrary to common belief, workers look for work." You don't say! One more time; people just want to do their best. My direct experience is that these very same people - the workers - are quite capable of self-directed improvement of their immediate environment - with the global objective in mind. I've experienced (and facilitated) the very worst section in a company as they changed themselves into the very best. Same exact people, same exact place, all they wanted was a little respect, half a chance, and an overall understanding of the end game.

[And it is worth noting here that the quickest way in the world to destroy improvement is to "move" some of the people who have improved to an area that won't improve. This punishes the improvers and rewards the non-improvers. Do nothing instead (oh there - it's a stopping action once more) and your self-improvers will continue to improve even more, and the non-improvers will realise they are standing still and that only they can help themselves.]

Okay, back to the topic. Did I say Roadrunner was important! So, what does it do?

Roadrunner, like kanban, functions to stop people from working when there is no work to do. It does that by having only two modes: (1) working at your normal comfortable productive pace when there is work to do, and (2) not working when there is no work to do. It overcomes the insidious and pernicious (natural) habit of slowing down and pacing according to the workload or lack thereof. Taylor had a term for this, he called it soldiering. Slowing down, soldiering, destroys capacity. Stopping on the other hand creates capacity. Just like andon does in fact.

If you don't do it, you will fail. (Critical Chain project management is exactly the same. Larry Leach to his credit mentions Roadrunner in Critical Chain Project Managment - I think that is the only time I've seen it mentioned in that context - Goldratt excepted.)

One more thing. As I was editing through, I was reminded of Goldratt's Rules of OPT, the second one of which is "Utilization and activation of a resource are not synonymous." The only way to make them synonymous is to de-activate non-constraining resources - which might be all of your resources - and only utilize them, when necessary, otherwise you will overproduce.

The Deep Irony

The deep irony here is in the stopping and its absolute importance. It's ironic because everyone thinks that these logistical solutions are about not-stopping; that they are about maximising the output of the system (and therefore all of the parts as well). And it is ironic because the developers and the practitioners know this - it's explicitly written down as I have just pointed out - it is no mystery - and yet it is completely ignored.

We see exactly this misunderstanding or lack of understanding in that brief extract of Womack and Jones about takt time and increasing output. We also see that in the West's interpretation of Ohno's wastes. Ohno was not removing waste per se; he was building capacity. How? Wait for it! By making things easier. Where did we hear that before? 1896 - Taylor and the pig iron experiments. You can only make work easier; you cannot make work harder.

Margaret Wheatley found the early literature frightening - because she did not have a clue about what it meant, and neither do the people who have told and retold and reworked the Taylor story or the Ohno story into their own mechanistic worldview ever since. The most important aspects of Taylorism, Ohno's TPS, and Goldratt's DBR are built around not doing things. That is, in stopping. As Brendan Fox once said in a comment or a post here on LinkedIn - and I am paraphrasing - operational problems are over-production problems. Where is the evidence of that over-production? Oh shucks; it's "stuck" as work in the process.

Remember Ohno said:

"The situation where "the parts were made" is surprisingly common.? Everyone worked hard and the parts were made.? If you asked me "What is the most important part of production control?" I would say it is to limit overproduction." (Taiichi Ohno's workplace management, pg. 36)

The moment you have this mindset, the value of stopping, then a whole lot of things begin to fall into place. Taylor, Ohno, Goldratt all knew this. Most everyone else does not. The paradox is to stop more often not less. The frequency will increase but the total duration will drop off dramatically and quickly; output will go up.

Not knowing this, not doing this, is the vice. To better understand the vice, let's first turn to the virtue.

The Virtue

All of these approaches have a virtue; and that is undeniable, they must have, they are exceedingly productive and exceedingly prosperous. So, let's start by drawing the virtuous part, and then we will return to, and incorporate the irony of the vice.

The diagram below is in essence the same diagram as we had in Part 1, and which was reproduced at the beginning of this page. Instead of being one of Shewhart's natural process charts, it now represents a physical system of some sort. A physical system with a constraint which is depicted at one end, the far right, for simplicity, and with flow from left to right. The constraint could be internal to the system, or it could be immediately adjacent to the system but be in the larger external environment or the market.

All of the logistical/operations systems that we have mentioned have raised output spectacularly with respect to any other approaches by exploiting something (Margaret Wheatley would both dislike that word and think it quite appropriate - because she thinks it means working harder). In fact, we achieve this by the virtue of succeeding to restrict the aetiology of exploiting everything everywhere all of time, to just a few things, maybe just one, all of the time. We, through someone else's thoughtfulness, restrict our own instinctfulness. And we are the better for it.

Thus, Taylor by focusing on human muscle power raised the output of his system - a single individual - to 47 tons of pig iron a day. How much physical exertion an individual man could sustainably do was the constraint.

Thus, Ohno raised the output of his system such that today it is the world's largest automobile manufacturer. He focused on the dependencies and variation in the parts that constrained his system improving capacity as he went. When he started in 1937 "it took nine Japanese to do the job of one American." (Ohno pg. 3)

Thus, Goldratt showed how, as a universal concept, the output of any such system is governed by a singular constraining step - somewhere, either internally or externally. But there is something more. In any system, as Deming said; you have the obligation of a component. I like the term subordination to describe that. Let's complete our diagram.

The Vice

Making use of someone else's codified thoughtfulness allows us to restrict our own instinctfulness with substantial benefit. That's the virtue, as we have said. The vice is when we then fail to further expand upon our own thoughtfulness. Let's have a look.

The vice, and indeed the deep irony, is that we then fail to apply the teleology of subordination to the few common things that cause most of the problems most of the time everywhere else; in fact, to the non-constraints (and regardless of, or quite independent of, any singular constraint and its exploitation). We don't expand our own thoughtfulness. Every stopping action in the examples above was a major subordination action of the non-constraining parts - maybe all the parts in fact - to the objective of the whole. Less speed more haste, or get there faster by going slower, or get there sooner by starting later. Our language has cliches to explain these things.

This is why TQC or TQM or Six Sigma, as partial interpretations of SPC/SPK, fail; we don't get the second part right. We stop tampering, which is good, but we don't start addressing the remaining variation which is bad. We fail to recognise the general or systemic nature of these things. OK, such approaches do succeed in keeping everyone busy everywhere all of the time, but prosperity does not increase.

This is why lean and agile, as partial interpretations of TPS/SPC, fail; we don't get the second part right. We run around reducing waste, which is good (kinda), but we don't start to address building capacity from it, which is bad. Neither do we learn how subordinate that capacity even if we were to build it. In fact, we often "let go" the very people whom we exhorted to improve. We don't learn.

This is why TOC fails; we don't get the second part right. We run around identifying and exploiting the constraint, which is good, but we don't start subordinating the non-constraints, which is bad. It is just like quality, we exploit the constraints with a few tactics that are specific; that is, they are local and deterministic in time and place and degree. However, we can't handle subordinating the non-constraints with a few tactics that are general; that is, global and stochastic in time and place and degree.

If you do subordinate your non-constraints, tell me about it in the comments.

And look, the beauty of this - the virtuous part - is that for every exploitation tactic there will be a complimentary subordination tactic; that's 1:1. That's singular cause and effect. Our brains can handle that. Equally, the beauty of the more thoughtful subordination of dominant general causes that afflict the non-constraints is that they are always a just a few too. Just a few, each with a one-to-many relationship - call it leverage if you like. Let's resort to metaphor; we can "kill" several hundred birds with just one stone. Well maybe 3-4 different stones. But it's still a lot of birds.

Like in our investigation of SPC/SPK where there are only a few dominant general causes, the same applies here. Thus, there are also only ever a few small critical injunctions, maybe 3-4 (rather than 4-5 as in quality), that act as countermeasures to these dominant general causes - and they apply equally to all of the non-constraints.

Even though the non-constraints might number in their many hundreds (say individual machining centres), the subordination tactics will always be a very few in number, however, they will apply equally and generally to all of the non-constraints. They are general in time and place and degree.

Not stopping is emblematic of this failure, this vice, it is also the simplest, easiest (you stop) and most powerful to rectify. The way that stopping is so counter to our instinctfulness simply stresses its importance even more. If I was to list out the few subordination tactics that I use to align the parts with the whole, you would find they are all "stopping" actions of some sort. We are trying to make work easier - remember? It is our instinctfulness that makes it harder!

If I twist a common enough saying in TOC around a bit, we get this: we can ignore the incorrect or non-existent subordination of the non-constraints - in fact, we usually do; but the non-subordination of the non-constraints will not ignore us.

Here's a thought. If there is an 80/20 rule operating in Theory of Constraints, exploiting the constraint gets 80% of the attention and yields 20% of the results, subordinating the non-constraints (if actually ever done) gets 20% of the attention and 80% of the results. Think about it.

The Apparent Contradiction

So, does that mean the constraint in Theory of Constraints isn't important? No, it does not. What Goldratt knew (Jim Cox has told me this more than once) is that 2-3 times around the internal process and the constraint will be in the market. That's kind of a home truth people don't want to know. They want to be "constrained" they want to be able to blame the lack of something on someone else for their current predicament. The reality is that we are always over-producing. How in the hell do we over produce if we are internally constrained? The answer is we are not internally constrained (other than by our aetiological policies).

When the internal constraint is broken you then have an external market constraint, your internal non-constraints must access or address that external market constraint through the few small critical paradoxical injunctions that are countermeasures to the 3-4 dominant general causes that account for most of the (lack of) current market effect. Somewhere there has to be a rate-limiting step, this was Goldratt's thesis. Except now it is external and by definition everything else internally must be a non-rate-limiting step, there is no other way. And so, the non-rate-limiting steps must fall back to the same limiting-rate. In production that is called stopping. There is no other way that doesn't destroy capacity and prosperity.

Aetiology, Teleology and Psychology

Deming's System of Profound Knowledge has four components (New Economics pg. 93).

• Appreciation for a system
• Knowledge about variation
• Theory of knowledge
• Psychology

Some people have a knowledge of variation, and some people have a knowledge of system - and for good measure here is the best definition of a system that I've ever seen:

"A system is a whole that is defined by its function in a larger system of which it is a part." (Ackoff, Differences that make a difference pg.119)

It is recursive - because systems are like that.

When it comes to Deming's third point, theory of knowledge, however, the "some" has thinned out to "a few." Bateson and Ackoff are good places to start. But by the time we get to psychology things have thinned out to hardly anyone at all!

I wondered if I was really in a position to address the last part, psychology. I went on a "Herbie Hunt" looking for my earliest exposure. It was two and half decades ago: Introducing NLP: psychological skills for understanding and influencing people. Sounds like Dale Carnegie stuff - right? Well, you've got to start somewhere. And what a good place to start. It featured Gregory Bateson, and he has featured ever since. The next was a little darker: Robert Hare's (1993) Without conscience: the disturbing world of the psychopaths among us. I'd worked for a few, I was learning. I've worked for a few more since, I can't seem to avoid them. All up I found works by 16 psychologists/psychiatrists, and few more by lawyers (Getting to Yes, Getting Past No, Difficult Conversations) and musicians (Fritz's The path of least resistance), and one by a psychiatrist/musician (Heifetz?Leadership without easy answers). But now I am telling you more than you need to know.

Let me bring you up to the present. I had the good fortune recently to have a colleague recommend to me two books. They are The Courage to be Disliked, and The Courage to be Happy, by Ichiro Kishimi and Fumitake Koga. A sort of "What happens when two Japanese men encounter ancient Greek philosophy and distill pieces of Stoic and Socratic brilliance with all the practical force of Adlerian psychology?" Adler, it turns out, has informed any number of other sources that I have come across: Berne's Games People Play: the psychology of human relationships and also Elliott Jaques of Requisite Organization fame to name but two.

Kishimi and Koga constantly juxtapose the aetiology of the student in their Socratic dialogue with the teleology of the master who is educating the student on the understanding of a greater master still - Adler. The student constantly expresses our natural aetiological view where a current situation is ascribed to something bad that happened in the past. The master constantly and skilfully rebuffs and reconstructs the arguments teleologically so as to ask what is the purpose of the bad effect that the student chooses to maintain by ascribing it to a past event. Essentially what is the benefit of maintaining his present situation by blaming it on something or someone in the past.

Now I have to admit that I thought, up to that point, that aetiology was the opposite of teleology - like atheist is to theist. One focused on input cause and one focused on output effect. But that's just my bad English and there is no accounting for that. Aetiology means the investigation or attribution of cause or a reason, and the root (attia) means from a cause. Teleology on the other hand means the explanation of phenomena by the purpose they serve rather than by postulated causes.

I was aware of teleology because it is natural to the natural sciences - and that is my training. It turns up in Bateson (Mind and Nature pg. 56) - remember I can't shake him off - and also Ackoff (Ackoff's Best pp. 22-24). Ackoff says of aetiology that it is a mechanistic or deterministic and input-oriented way of looking at things, - aka causes, whereas teleology allows us to look at things in an output-oriented way.

"Objective teleology does not replace determinism, which is an objective ateleology [ah! that's the source of my confusion]; it complements it. These are different views of the same thing, but the teleological approach is more fruitful when applied to systems." Let's continue with Ackoff just a little more: "In the Machine Age, even humans were thought of as machines. In the Systems Age, even machines are thought of as parts of purposeful systems."

Let me pause here to say that it seems to me that people in the "hard" sciences seem to use aetiology; their effects, after all, are often without purpose. Whereas people in the "field" sciences and social sciences use teleology more; they aren't dealing with effects so much as the outcome of effects, the purpose, a higher logical level no less.

The reason my colleague recommended Kishimi and Koga to me is that it resonated with him as I related, via e-mail, chapter and verse two recent books that I had read: Change: principles of problem formation and problem resolution by Watzlawick, Weakland, and Fisch, and then The Tactics of Change: doing therapy briefly by Fisch, Weakland, and Segal. The latter especially was ground-breaking in its rejection of psychotherapy's penchant for long and prolonged (read financially rewarding) delving into people's past in search of an appropriate [recovered] cause for their current behaviour. A cause that, for god's sake, doesn't exist anymore; except in a real or imagined [reconstructed] memory (which might not have even existed until the psychologist started to rummage around). Instead, they sought to understand the positive good, the purpose served by the maintenance of bad effect (the "psychotic" behaviour). Moreover, once having identified that, they dealt with it using brief therapy built around paradoxical injunctions. For the agoraphobic they might suggest, "don't drive too far from home" as a very simplistic but very real example.

But there is more than that. Firstly, they were using teleology in place of aetiology - although I don't think they ever made either concept explicit - and they were all juniors to Gregory Bateson- although they didn't mention that either! Bateson had planned a book on schizophrenia but was pipped to the post in 1966 by Pragmatics of Human Communication by Watzlawick, Beavin and Jackson. Not only did Bateson not know of this - and it drew directly upon his own work - but insult was added to injury when it was his own publisher published that book behind his back, having already delayed negotiation over his own book on the matter. (Harries-Jones A Recursive Vision: ecological understanding and Gregory Bateson pg. 27-29).

Okay, let's try and tidy-up a few threads before heading into the home stretch. Firstly, my colleague recognised the implicit aetiology and teleology of Watzlawick and Co., through my descriptions and absolutely nailed it back to Adler via Kishimi and Koga where it is made explicit. In both sets of writings brief therapy and paradoxical injunctions are paramount. Thirdly, I had grown to use in operations/organizations the concept that we can fully align the parts with the whole through just a few, usually 3-4, simple paradoxical injunctions. There is a congruence here.

Ohno's Logistics and Thinking are Teleological

We discussed much earlier Ohno's observation that "the parts were made," but he continues in the very next essay to ascribe this to an agricultural heritage. That heritage is a cause for the effect of over-production. But there is also a purpose expressed as well: security, "We still behave as if we are farmers (Workplace Management pg. 37). Although he made a distinction between former hunters and later farmers, this story, this narrative, regardless of its ultimate validity is an explanation at its essence about our instinctfulness. It is why we do something; it is teleological in its intent.

Now let's juxtapose that with a common Western interpretation of TPS. The Western interpretation of reducing waste is an aetiological interpretation, a "what." It goes no further than that. The "why" however, the teleological interpretation, is a higher logical level purpose, which is to create additional capacity, addition productivity in fact - by making things easier. Increased productivity that produces increased prosperity.

And of course, the purpose of Toyota Motors wasn't to reduce waste, it was to produce motorcars. We seem to forget that. The Toyoda's sold up shop on Toyota Spinning and Weaving because they were obsessed with building motorcars. And for that matter Toyota Spinning and Weaving came about because they were obsessed with mechanising hand-weaving (in general people have almost no conception of the amount of time embedded in hand spinning and hand weaving that was required in temperate and cooler climates prior to mechanisation.)

Lastly, "A person standing in a production area can end up cleaning the corner of an enormous box with a toothbrush.? Toyoda Sakichi and Toyoda Kiichiro were different and always studied the entire picture.? They had the overview of chess players and were constantly designing strategies.? They knew how to checkmate." (Toyota Production System pg. 86) Now be careful, Japanese are always self-effacing. You always get the message in a story about someone else. The outlook expressed here is systemic, it is purposeful, it is outward in its intent. It is certainly not deterministic; it is certainly not or mechanistic.

Now, you can accuse me of cherry-picking a couple of things. But it's really up to you whether you want to understand the point or not.

Goldratt's Logistics and Thinking Process are Aetiological

Goldratt's logistical processes are systemic, that cannot be denied. But that, ironically, is an outcome or a consequence of the application of a mechanistic and deterministic OR application called Linear Programming. Linear Programming will always produce a most-constrained through to a least-constrained grouping. We know that Goldratt first assumed several interactive constraints within his OPT scheduling approach before settling on the notion that there can only be one (or interactive chaos), but non-constraints are always a consequence of the LP algorithm - and they have "freedom."

Goldratt's logistics and thinking process are both deeply aetiological. Even though the logistical applications are systemic, their overall intent is deterministic. They are developed around a deterministic and mechanistic singularity of a constraint. But it also goes deeper than that.

In What is the Thing Called Theory of Constraints and other places such as the introduction to The Goal, and in Essays on the Theory of Constraints Goldratt championed a classical scientific developmental sequence of: classification, correlation, and causation. Briefly illustrated with snippets from astronomy and other sciences - including genetics and microbiology.

But his point was always to get to the concept of causation and effect-cause-effect. This stage, the stage of seeking causality, was, he considered, the point at which a subject became a science. Let me redraw effect-cause-effect in two-dimensions. Thus:

Effect 1 Effect 2

\/

Cause

Goldratt's contention was that once an underlying cause had been postulated for an existing effect (effect 1 here) it could then be utilised, in fact must be utilized, to predict another effect (effect 2 here) that was either previously in existence and not understood or correlated, or it was not even in existence (or rather awareness) and subsequently discovered as a consequence.

Deming would simply call this evidence of a theory.

"The theory of knowledge teaches us that a statement, if it conveys knowledge, predicts future outcome, with the risk of being wrong, and that it fits without failure observations of the past.

Rational prediction requires theory and builds knowledge through systemic revision and extension of theory based on comparison of prediction with observation." (Deming pg. 102)

I would just add here the notion that the more concise a theory is, the simpler it is in fact, the more than it must therefore encompass, and thus the greater risk that it could be wrong -which paradoxically reinforces our own belief in its correctness. Put that aside.

We might do a small thought experiment with Theory of Constraints as an example:

Rate-limited system (1) Non-rate-limited parts - non-constraints (2)

\/

Rate-limited part - constraint

LinkedIn will probably collapse my spacing. Reconstruct it in your mind.

Using effect-cause-effect we might postulate that a rate-limited system is caused by a singular (or just a few) rate-limited parts. We would have to then predict that as a logical consequence there will also be non-rate limited parts (perhaps very many, if not all).

Now I put the word logical in there on purpose. Because Goldratt overstates the argument about logic. "All our inventions/decisions/convictions are based only on intuition (the communication of this to ourselves and others is based on logic.)" (Goldratt pg. 79) This is not so. I think this is a small but significant point; you can also communicate tacitly. I can teach the principles of any of the logistical solutions tacitly - without saying a word. Okay, just put that aside too.

The whole purpose of this little exploration was that Goldratt's physics taught him to see for any effect, an underlying cause. That is aetiology. Even when discussing the genetics of Mendel, or the microbiology of Leeuwenhoek/Pasteur, Goldratt keeps his argument aetiological; deterministic, mechanistic, cause and effect. But these latter two examples are also purposeful, as is everything we do with operations/logistics/quality.

As per Ackoff's comments earlier. Even a machine or a whole process can have a purpose in a larger purposeful system. But Goldratt seemed to take a very mechanistic and deterministic route even though the system and the people in that system are manifestly purposeful.

The real question isn't why the virtue, the real question is why the vice?

The real question isn't why we must exploit the constraint; the real question is why we don't subordinate the non-constraints.?

The real question isn't why we look for special signals outside of the noise, the real question is why we don't look for general signals in the noise?

What is the purpose (virtue) of separating the signal from the noise? You might say it is to ensure we take specific actions only when and where they are warranted to remove a specific cause. In essence it is to limit our instinctfulness.

What then is the purpose (vice) of failing to investigate the signals that still remain in the noise? To protect us from our insecurities that arise from actions dealing with general causes. In essence to protect our instinctfulness.

What is the purpose (virtue) of exploiting the constraint? You might say it is to ensure we take specific actions only when and where they are warranted to increase production. In essence it is to limit our instinctfulness.

What then is the purpose (vice) of failing to subordinate the non-constraints? You might say, once again, to protect us from our insecurities of taking general actions that might apparently lead to under-production. In essence to protect our instinctfulness.

Doing what seems to preserve our (instinctful) securities is purposeful. Not doing what seems to not preserve our (instinctful) securities is also purposeful.

Summary

Do you see the irony in Taylor? Many people wish to believe that he was making work harder, whereas he was making work easier. That goes against our instinctfulness, but is congruent with our thoughtfulness. Prosperity rises.

Do you see the irony in Toyota? Many people wish to believe that they were reducing waste, whereas they were limiting over-production and building capacity. That goes against our instinctfulness but is congruent with our thoughtfulness. Prosperity rises

Do you see the irony of Goldratt? Many people wish to believe that he is exploiting the constraint, whereas he was subordinating the non-constraints. That goes against our instinctfulness but is congruent with our thoughtfulness. Prosperity rises.

There is a finer nuance, however. Both the vice of quality and the vice of operations/logistics are general causes. Our aetiology biases us towards the specific: the local and deterministic in time and place and degree. We find it difficult to address - at least without thought - the general: the global and stochastic in time and place and degree.

Yet these few general things, either in quality, or in operations/logistic, account for the vast majority of the observed effects.

Let's pull our simple diagram back one more time.

Restricting ourselves to focusing on the efficiency of the constraint rather than everything else everywhere is like focusing on the special cause rather than the noise.

Failing to subordinate the non-constraints with 3-4 simple paradoxical injunctions is like failing to investigate the 4-5 dominant general causes in the noise.

Let's close out at the same place that we came in at; in quality the real signal is in the noise, and in operations logistics the real productivity and prosperity is in the non-constraints. The story is exactly the same one.

Reducing tampering is one thing, actually improving quality is entirely another. Focusing on the constraint is one thing, actually improving the overall process is entirely another.?

We desperately want to believe that effects have specific causes: causes that are local and deterministic in time and place and degree. We have a very hard time to believe that the most important effects are, in fact, from general causes: causes that are global and stochastic in time and place and degree - even though just a few of these, maybe 4-5 account for the vast majority of the observed effects.

www.dbrmfg.co.nz