Showing posts with label assembly line thinking. Show all posts
Showing posts with label assembly line thinking. Show all posts

Saturday, November 17, 2007

I've been framed!


There are two ways to change the meaning of something - you can change the something, or you can change the context in which you say it. If we don't account for this, we will make terrible mistakes in communicating with each other, and even with ourselves. If we grasp this, we can overcome many of the problems that plague our world today, which are results of unrealized context shifts. We have content processors but what we need now are "context-processors."

We all know that a quote, taken "by itself" out of context can be totally different than what it meant at the time. This is often visible in courtroom dramas, where the person is asked by the attorney, "Answer, yes or no, did you say this?" followed by some damning phrase or sentence that sounds totally wrong out of context. We all know this is unfair and somehow wrong, but don't have a strong way to assert that or to understand how pervasive this effect is.

It doesn't just affect communications. It affects our ability to work alone!

My favorite expression of this truth was a cartoon one day by Charles Schultz of Snoopy, the dog, lying on top of his doghouse, staring at the stars and pondering. He said:
Did you ever notice
that if you think about something at 2 AM
and then again at noon the next day
you get two different answers?
This cartoon is profound. Slow down and consider that this means. This says that a correctly functioning human being has a context-sensitive thinker-thingie that produces different answers to the same inputs depending on what larger context it is sitting in at the time.

This is, in my mind, a "feature not a bug." In fact, this seems to me to be the key to reconciling humans and resolving age old conflicts that have seemed totally impossible to tackle.

This is also a critical insight in trying to figure out how to make decisions today that don't seem totally stupid tomorrow.

That's true whether you are a person, a group, a corporation, or a nation.

We are walking around comparing "content" and failing to account for different "context" in which that content was perceived or generated. In small, local worlds where context is shared and identical among people, we used to be able to get away with that. Once we start trying to cross cultures or "silos" of expertise, and do something interdisciplinary or international, this tends to trip us up every time. We didn't learn the "general case."

Content is explicit, obvious, the kind of thing you can hit with a hammer. Context is implicit, invisible, unstated, and hard to describe even when you try. But it is vital that we learn how to do this, to get by in a diverse world - a world in which different people are operating in different contexts but trying to communicate with each other over space and time.

It is crucial when we try to take some thought or observation, about a patient, say, and "record it" in some electronic database where we will pull it up a year later and compare the two to see what changed. Are we capturing what we need to do that assessment correctly? Are we writing something down in words that will bring up the right thoughts to a different doctor next year?

Tragically, we have failed, socially, to understand the full implications of this issue. The miracle of technology allows us to store or send content across space at the speed of light, but, oopsie, forgot about the context part of the message. What gets delivered is not what was sent, in huge ways.
It does not have to be this way. In the same way that we have built computers that do content-processing correctly, we can build environments that do context-processing correctly. It is critically important that we learn how to do that.
Now, these effects are not flaws in humans that would go away if we were all "rational" or "scientists" or if we all based our judgment on "data" and "evidence." These effects are properties of the very nature of space, time, and information itself. We cannot "get around them" or ignore them. We are going to have to learn how to account for them correctly.

It doesn't have to be hard, but it does have to be done, or we'll keep fighting needless wars, between parties that actually agree with each other but don't realize it.

Take the example of "perspective" -- a distortion of space where it appears to each observer that things "far away from them" are small, and things "close to them" are large, and as you move towards a distant building or mountain it "gets larger."

At some point in life as infants we figure out that the thing we're looking at actually isn't changing size at all, it's an illusion, a distortion, caused by where we are looking from, our viewpoint. If we didn't correct for this, we could argue all day about which of two things was "bigger" and what was "fair" and not get a resolution, because A looks bigger to me than B, but looks smaller to you. Once we correct for that perspective distortion, we can resolve that question in a way that makes us both happy. This happened so early in our lives we forget we had to learn it.

There is a popular misconception that because things are "relative", there is no underlying reality, and no way to ever reconcile them. Einstein said the opposite. He said that actually, once you understand what is going on, you can completely reconcile observations made by two competent observers, relative to their own reference frames, all the time, every time. You can totally account for the changes, say, in perspective between two observers, and figure out entirely how the world I see needs to be warped and twisted to give the world you see.

Computer animators and virtual worlds have to deal with this "perspective" or "viewpoint" transformation all the time. It's a lot of bookkeeping under the covers, but straight-forward if you do it carefully.

Unfortunately, there are other shifts in context that are less familiar to us that impact our ability to reach agreement. The problem is very deep, as I said, built into the nature of space and time itself.

Once, in my astrophysics grad student days, I took a course in Cosmology and General Relativity. I didn't get it all, but I got enough to get the story, which is not that hard to tell, and does not require math. Please don't flee. There will be no quiz. Everyone one here gets an "A".

The essence of Einstein's General Theory of Relativity could be distilled down to three insights, widely misunderstood. You don't need to be Einstein to understand these ideas.

The insights are that
  • content and context are completely equivalent in what can be said, though widely different in what can be expressed easily in each;
  • you cannot change one without affecting the other;
  • if you do your sums right, the changes are completely predictable.

In the world described by relativity, the meaning of a very concrete phrase or physical expression or very real measurement of velocity, say, changes as you slide it around in space and time, most famously if you attach the observation to different observers traveling at different speeds. I can only observe your speed "relative to me", so depending on how fast I'm going, I'll measure something different.

This is no big deal. People in a car on the highway appear stationary to each other, even though the car is speeding down the road.

Or, if I'm standing on the Earth, I see the sun, obviously moving across the sky, going around the Earth. If I'm standing in some space ship off to the side, I see the earth spinning and the sun remaining fixed. These observations are both right, relative to the "reference frame" in which they were made. To reconcile them you have to account for the different in frames used by two completely competent observers.

Implications
==========

Well, what does this mean? For one thing, it means that where you work or spend time thinking or talking to each other affects what result you'll come up with. A decision that is "obvious", made in a bunker-like dimly-lit War-Room deep underground might not be at all the same decision that would have been made, given the "same facts", in a cheery, sunny deck in a woodsy retreat on a warm spring day.

It means that an "obvious" decision about what to do next, made standing in an urban war zone with explosions in the distance is not the same decision, given the same information, that is "obvious" viewed by people safely out of harm's way, at their leisure, later, reviewing the tapes over coffee and some nice Danish.

In fact, as a child, I observed that the behaviors that seemed to make the great leaders "great" in war movies wasn't that they were brilliant, but that they simply managed to remain stable and sane when the world around them had gone to hell. They remained connected to a larger, stable world despite the fact that their body was located in a locally unstable one.

Maybe, there is value in having content-intensive work like "science" embedded in larger stabler social frameworks that religions have sometimes produced in the past. I find it fascinating that, according to a recent issue of New Scientist, geneticists are discovering that far from being "junk DNA", the DNA between the 22,000 genes that code for proteins (content) may be even more important, and this "junk" codes for the larger context that decides when and whether that content should be expressed, or modified in the way it is expressed. (Junking the Junk DNA, New Scientist, July 11, 2007).

Yesterday, I did a post on the software world "Second Life" and possible roles of virtual reality in getting people to experience worlds they couldn't get to on their own. Today, I want to add to that the idea that virtual worlds are virtual contexts, which means that you can conceivably adjust not just the contents of a scene, but the context of the scene in which those contents are embedded.

This may be the tool we need to explore more how context and content interact with each other for humans, and to learn how susceptible we all are to "framing" of an issue. We can understand how advertisers or demagogues try to use propaganda techniques to shift the frames of discussions so that, even though we seem to be the same people, we end up making different decisions. Even though we don't feel manipulated, we have been - by Madison Avenue agencies that know how to send broadband messages in context-modulation that bypass all our cognitive protections against content-manipulation. That's what TV is all about, to them.

Dirty pool aside, honest and diligent CEO's and civic leaders need to understand what an idea will sound like, or be taken to mean, in hundreds of different contexts, to know how to process the input they get, or how to say anything that won't offend one group or another.

If nothing else, cars for Latin America, for example, shouldn't be named "Nova" - since "No va" in Spanish means "won't go." Underarm deodorant shouldn't be advertised in Tokyo using a happy octopus logo, since in Japan an octopus doesn't have 8 arms -- it has 8 legs. Oopsie.

(photo Walking Alone, by me, on Flickr)

Thursday, August 24, 2006

Systems thinking and health care - reflection

Thursday, August 24, 2006

Systems thinking about assembly line health care

Why are the Big Three auto companies going broke, and if their high-tech Detroit area assembly-line operation is our model for health care, is that our destiny as well?

Consider this recent news.

New York Times
August 19, 2006
Ford Is Slashing Production 20% for 4th Quarter
By MICHELINE MAYNARD
DEARBORN, Mich., Aug. 18 — The Ford Motor Company, which is struggling to keep its grip on second place in the American car market, said Friday that it would cut by one-fifth the number of vehicles it plans to build in the final three months of the year.

The slowdown represented the deepest production cuts since the industry’s crisis of the 1980’s. It also underscored the difficulty that Detroit, whose business relies on sales of sport utility vehicles and pickup trucks, is having as gas prices remain around $3 a gallon. Detroit’s market share has dropped to its lowest level in history, while Asian brands, known for their fuel efficiency, are setting sales records...

On Friday, Ford officials contended that no one in the industry could have anticipated that gasoline prices would remain so high ... [emphasis added]

The mental-model of fine-tuning assembly-line clinical medicine may break down when seen through public health's wider lens, or in light of the impact on the humans whose bodies and lives provide the pixels on which this video is displayed, or be fine in both those contexts but have issues when seen in terms of globalization and the increasing ability of patients to fly to Singapore for surgical care at a total cost well less than similar care here.

The "systems thinking" approach suggests that we should spin through all those lenses on our microscope and consider whether each one makes sense, before we fixate, as Ford and GM seem to have, on activities that are perfectly sensible when removed from global context, but that look different when viewed in situ.

In clinical care systems we know that the parts do not exactly fit, and are not always well-coordinated. We take as an unexamined given that with just a little better training, a few more messages here and a new EHR message-store component there, surely this will come-together, click, and emerge as a dynamically stable and scalable system that cranks out top-notch patient care and dollars like a fine-tuned machine. Won't it?

In the 21st century, Peter Senge said, it is not enough that organizations persist well, they also must be able to change well. And Cybernetics 101 teaches us that in order to change well, a system needs to be able to see well. Perception is everything. We should confirm that those specs are reflected in the Electronic Health Record brave new world we're building nationally, before too much concrete is poured that will need to be ripped out and redone.

Like GM, we both face the twin questions of how humans fit into the assembly line model, over extended time periods, and, stepping back, whether the product those lines are producing is, viewed in global context, the product consumers will be buying next year.

We have to make sure not only that we're solving the problem right, but that we're solving the right problem. And, in the way of systems, we have to ask whether the way we're structuring humans and machines to solve the one problem doesn't create, as an "unintended side-effect", an emergent interference with our ability to see many small signs on the front lines that things are changing and the old model doesn't really fit the new reality.

Or are we somehow blinding ourselves to marketplace changes that seem obvious to the world not embedded in our own internal story? Ford executives asked who could have known that gas prices might rise? Hmm. Somehow they managed to assemble a workforce of half a million people with a resultant net organizational IQ of less than the average teenager. We need to ask where their collective common-sense-emerging process broke down, to be sure we're not following in their footsteps and mimicking their behaviors a little too closely.

Obviously, the problem is subtle. This is the way with most system problems.

There's lots of blame being spread around for "obvious" events and causes and decisions, including health care costs in another reflexive feedback loop, but we of systems thinking world know those are typically red-herrings and just symptoms of far more subtle problems in systems dynamics space. We are in desperate need of a "macroscope" that makes this almost invisible tissue stained and visible to the naked eye at last so we can see not just the microbes that we deal with daily, but the living dysfunctions that appear capable of infecting human organizations and causing Detroit blindness.

I'm not forecasting what tomorrows customer and problems will be, only that they will not be the same as yesterday's, in some really important ways. But I am saying that, based on the inexorable evidence and outcomes of the US automobile industry, there is something deeply but subtly wrong with the assembly-line model on a system dynamics basis - something that results in organizational blindness and incapacity to adapt.

It would seem the better part of valor to figure out exactly what that is, and make sure we don't inadvertently copy that into health care delivery systems as well.

Bottom line, literally: it would seem worth easily a billon dollars and a million votes to the Michigan economy to understand that answer in depth, and not just superficially. Whatever it is, there's a lot more where it came from.