![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
006: Organic v. Mechanical Metaphors
Apr. 14th, 2021 02:28 pmThink of a system. What picture comes to mind? A computer, or a roomful of them? A decision tree, with questions and arrows branching to different answers, all of which converge on “Then don’t worry”? What about a lake, the topography that allows it to stand, the rainfall, the vegetation, the things that live in an around it and eat one another? A system to Webster is as simple as a bunch of interconnected parts that form a whole. All of those examples qualify, and many more. What I want to explore is whether we can run into trouble by comparing certain types of systems that aren’t really comparable. A map is a metaphor, and all metaphors break down eventually, but eventually is a lot better than right away for the same reason that I don’t want to pull a new map out of the glove compartment every block. I’d rather a map with mileage.
Let’s consider a type of system that a lot of people like to write books and internet articles about: a business. Is it better for a business to run like a well-oiled machine, or a gazelle? Who cares, either one. Both run fine. But how do you handle them when they start to run poorly? What are the possible causes? In the machine metaphor, it’s probably the wearing down of a part, which you handle by replacing it. Turn it off, identify the part, get a new one, turn it on. With the gazelle, it isn’t so simple. You can’t just pop in a new rear left leg because the old one is a bit gimpy. It would need rest, food, water, and protection from predators while the healing process worked itself out. You would understand that muscles and organs aren’t just interchangeable, and that invasive surgery would have system-wide consequences, at least temporarily.
In this case, maybe a mechanical metaphor really would lead to better decisions (certainly easier ones), but it’s obvious that the choice of metaphor changes the context for the possible problems and solutions we identify.
Characteristics of Mechanical Systems
AKA machines. A machine is made of inorganic parts. (I know there might technically be one somewhere that uses a very specific organic component, but for the most part, it’s inorganic and non-living.) As of this writing, Man is the only species that makes machines. We don’t wake up and find them immaculately delivered to our doorsteps, which is to say, they are made by a person, and someone, somewhere, understands every part and purpose of the system—in fact, he designed it. If something goes wrong, a repairman who doesn’t live inside the machine is called, he opens it up, and fixes it. The parts are interchangeable, instantly replaced by a duplicate. The types of inputs, outputs, processes, etc. are determined in top-down fashion by the designer. It’s unlikely that a purely mechanical system will decide on its own to reorient its parts. Some computer systems maybe make changes to themselves, but those changes occur according to sets of rules given by, again, the designer.
Machines follow the rules—whether the way the parts are arranged, or a software program—to victory or death. They don’t seek reward or avoid pain. Most don’t learn a thing from their failures. Some might be capable of simple learning, what Bateson calls Learning I, which is memorizing new information and incorporating it into future actions. Very few computer programs love failure, in that they make a large number of low-commitment exploratory decisions to gain information that they might use to better direct future actions. When they do, they can only do so in the narrow framework provided by the programmer, which means contextual biases are built in. Most mechanical systems just return an error code, get stuck in an infinite loop, or in the case of the robot vacuum cleaner at my gym, wedge themselves under the same squat rack ten minutes into their cleaning cycle every single day.
In other words, they tend to be fragile, or at best, robust, but not antifragile.
Characteristics of Organic Systems
Organic systems, like a human being, are made of organic materials through continuous organic processes. Rather than being made, delivered, and repaired later if damaged, organic systems are self-healing, and these processes are constantly in motion. Some damaged parts regenerate, others do not. Often when that’s the case, accommodation is made around the damage so that the system can continue to function. To what purpose? There are no shortage of claims, but we can never know short of knowing the mind of God, or Nature. The design is obscure, and frankly, so is the function. As much as we can know about the human body (which is only one small example of an organic system), there remains a vast ocean of unknowns and disagreements about how to interpret the known. The processes are bottom-up. They happen without central direction, and only when things go really well or really poorly does it ever come to the boss’s attention. Organic systems don’t choose what types of input they’ll take. Humans are stuck with the five senses and sense ranges we were lucky enough to be born with. Presumably these were chosen via selection pressures.
Organic systems tend to seek rewards and avoid pain. Sticking with the human example, we are capable of Learning I, but we are also capable of at least Learning II—learning about learning, the context in which it takes place—and arguably more. We may seek low-order failures to guide future high-order success, so between that and the ability to become stronger after healing from certain stresses, we and most organic systems are antifragile to some degree. While machines are entirely logical, organic systems never use Western logic with the exception of a handful of human beings on a handful of occasions. That leads to a different array of possible decisions. It might not be entirely accurate to say these decisions are made through some combination of intuition and trial-and-error, but that’s close enough to highlight the differences.
Those differences are nearly across the board. That means using mechanical metaphors to describe organic systems leads to—and in fact, ensures—significant errors in interpreting and predicting the behavior of the system. “Adaptation” is exchanged for “repair”. Anyone with more than a few weeks of work under their belt knows that when one person in a given position is fired or quits, and they’re replaced by a similar person in the same position, the working dynamic changes completely. Mechanically speaking, we would expect a new part to plug in and play exactly like the old. But organically, every part of the new system has to reorient itself to every other part and achieve a new, if different, homeostasis. When parts change, they are rarely even similar. They could be geniuses, or idiots. Even if no parts change, parts can change their minds. They can lose them.
Small, highly-skilled units formed for a specific purpose can veer, or turn 90-degrees in spite of the intentions of all members and directors. Remember that a boss exists on the same logical type as an employee. Even if his top-down directives are followed to the letter by everyone, the Company as the next-highest logical type is subtly doing its own thing through bottom-up organic processes based on selection pressures from both within, and the wider world.
Simple problems in machines become far more complicated in organic systems. In a machine, waste and inefficiency should be minimized, period. That’s often true of organic systems, but not always. It might look like the right move to the accounting department to layoff a portion of the workforce, but if it creates fear and resentment among those who remain, productivity drops, and the most-skilled employees who just lost their lunch buddy start looking for better opportunities at companies that are more stable, the ultimate cost can far exceed the waste of a few salaries. Or it might not, depending on the context.
The point is that framing organic systems, and their problems and solutions, in your mind using a mechanical metaphor can potentially lead to disastrous decisions due to a fundamental misunderstanding of how that system behaves. So what counts as an organic system?
1. Any system containing an organic part—something that is alive. A lawnmower is a mechanical system, but Larry’s Lawnmower Repair, with its single owner-operator, is an organic system. A city contains people, and thus is organic. The soil contains trillions of microorganisms. Also organic. So on and so forth.
2. Any system not containing explicitly organic material that nevertheless behaves organically. For example, “the economy” is an abstract system, not a living thing, but it behaves like an organic system because it’s nested in the higher order of “the human species,” and doesn’t exist without it. That’s why people fiddling with interest rates based on algorithms leads to unintended consequences. In fact, most things that are clearly mechanical systems become parts in organic systems when the system is considered a single component of something one order higher, as in the example of the lawnmower sitting in Larry’s Lawnmower Repair.
It might seem trivial to fuss about what kind of metaphors we use to examine problems, but it’s clear that whatever map we choose strongly biases the things we look for and don’t look for in a given territory. That wouldn’t be a problem given the capacity to cross-reference other maps, but it’s also true that some people confuse the one and only map they have at their disposal with the territory, and utterly fail to consider anything that falls outside its grid. We’ll have a better lay of the land if we keep a number of maps handy, use the mechanical ones for machines, the organic ones for the organic, and avoid confusing a map of one logical type with a system that works on a logical type of a higher order. The operational status of your Ford F-150 is not a reliable indicator of the health of the Ford Motor Company, which is not an indication of the health of the city of Detroit.
Let’s consider a type of system that a lot of people like to write books and internet articles about: a business. Is it better for a business to run like a well-oiled machine, or a gazelle? Who cares, either one. Both run fine. But how do you handle them when they start to run poorly? What are the possible causes? In the machine metaphor, it’s probably the wearing down of a part, which you handle by replacing it. Turn it off, identify the part, get a new one, turn it on. With the gazelle, it isn’t so simple. You can’t just pop in a new rear left leg because the old one is a bit gimpy. It would need rest, food, water, and protection from predators while the healing process worked itself out. You would understand that muscles and organs aren’t just interchangeable, and that invasive surgery would have system-wide consequences, at least temporarily.
In this case, maybe a mechanical metaphor really would lead to better decisions (certainly easier ones), but it’s obvious that the choice of metaphor changes the context for the possible problems and solutions we identify.
Characteristics of Mechanical Systems
AKA machines. A machine is made of inorganic parts. (I know there might technically be one somewhere that uses a very specific organic component, but for the most part, it’s inorganic and non-living.) As of this writing, Man is the only species that makes machines. We don’t wake up and find them immaculately delivered to our doorsteps, which is to say, they are made by a person, and someone, somewhere, understands every part and purpose of the system—in fact, he designed it. If something goes wrong, a repairman who doesn’t live inside the machine is called, he opens it up, and fixes it. The parts are interchangeable, instantly replaced by a duplicate. The types of inputs, outputs, processes, etc. are determined in top-down fashion by the designer. It’s unlikely that a purely mechanical system will decide on its own to reorient its parts. Some computer systems maybe make changes to themselves, but those changes occur according to sets of rules given by, again, the designer.
Machines follow the rules—whether the way the parts are arranged, or a software program—to victory or death. They don’t seek reward or avoid pain. Most don’t learn a thing from their failures. Some might be capable of simple learning, what Bateson calls Learning I, which is memorizing new information and incorporating it into future actions. Very few computer programs love failure, in that they make a large number of low-commitment exploratory decisions to gain information that they might use to better direct future actions. When they do, they can only do so in the narrow framework provided by the programmer, which means contextual biases are built in. Most mechanical systems just return an error code, get stuck in an infinite loop, or in the case of the robot vacuum cleaner at my gym, wedge themselves under the same squat rack ten minutes into their cleaning cycle every single day.
In other words, they tend to be fragile, or at best, robust, but not antifragile.
Characteristics of Organic Systems
Organic systems, like a human being, are made of organic materials through continuous organic processes. Rather than being made, delivered, and repaired later if damaged, organic systems are self-healing, and these processes are constantly in motion. Some damaged parts regenerate, others do not. Often when that’s the case, accommodation is made around the damage so that the system can continue to function. To what purpose? There are no shortage of claims, but we can never know short of knowing the mind of God, or Nature. The design is obscure, and frankly, so is the function. As much as we can know about the human body (which is only one small example of an organic system), there remains a vast ocean of unknowns and disagreements about how to interpret the known. The processes are bottom-up. They happen without central direction, and only when things go really well or really poorly does it ever come to the boss’s attention. Organic systems don’t choose what types of input they’ll take. Humans are stuck with the five senses and sense ranges we were lucky enough to be born with. Presumably these were chosen via selection pressures.
Organic systems tend to seek rewards and avoid pain. Sticking with the human example, we are capable of Learning I, but we are also capable of at least Learning II—learning about learning, the context in which it takes place—and arguably more. We may seek low-order failures to guide future high-order success, so between that and the ability to become stronger after healing from certain stresses, we and most organic systems are antifragile to some degree. While machines are entirely logical, organic systems never use Western logic with the exception of a handful of human beings on a handful of occasions. That leads to a different array of possible decisions. It might not be entirely accurate to say these decisions are made through some combination of intuition and trial-and-error, but that’s close enough to highlight the differences.
Those differences are nearly across the board. That means using mechanical metaphors to describe organic systems leads to—and in fact, ensures—significant errors in interpreting and predicting the behavior of the system. “Adaptation” is exchanged for “repair”. Anyone with more than a few weeks of work under their belt knows that when one person in a given position is fired or quits, and they’re replaced by a similar person in the same position, the working dynamic changes completely. Mechanically speaking, we would expect a new part to plug in and play exactly like the old. But organically, every part of the new system has to reorient itself to every other part and achieve a new, if different, homeostasis. When parts change, they are rarely even similar. They could be geniuses, or idiots. Even if no parts change, parts can change their minds. They can lose them.
Small, highly-skilled units formed for a specific purpose can veer, or turn 90-degrees in spite of the intentions of all members and directors. Remember that a boss exists on the same logical type as an employee. Even if his top-down directives are followed to the letter by everyone, the Company as the next-highest logical type is subtly doing its own thing through bottom-up organic processes based on selection pressures from both within, and the wider world.
Simple problems in machines become far more complicated in organic systems. In a machine, waste and inefficiency should be minimized, period. That’s often true of organic systems, but not always. It might look like the right move to the accounting department to layoff a portion of the workforce, but if it creates fear and resentment among those who remain, productivity drops, and the most-skilled employees who just lost their lunch buddy start looking for better opportunities at companies that are more stable, the ultimate cost can far exceed the waste of a few salaries. Or it might not, depending on the context.
The point is that framing organic systems, and their problems and solutions, in your mind using a mechanical metaphor can potentially lead to disastrous decisions due to a fundamental misunderstanding of how that system behaves. So what counts as an organic system?
1. Any system containing an organic part—something that is alive. A lawnmower is a mechanical system, but Larry’s Lawnmower Repair, with its single owner-operator, is an organic system. A city contains people, and thus is organic. The soil contains trillions of microorganisms. Also organic. So on and so forth.
2. Any system not containing explicitly organic material that nevertheless behaves organically. For example, “the economy” is an abstract system, not a living thing, but it behaves like an organic system because it’s nested in the higher order of “the human species,” and doesn’t exist without it. That’s why people fiddling with interest rates based on algorithms leads to unintended consequences. In fact, most things that are clearly mechanical systems become parts in organic systems when the system is considered a single component of something one order higher, as in the example of the lawnmower sitting in Larry’s Lawnmower Repair.
It might seem trivial to fuss about what kind of metaphors we use to examine problems, but it’s clear that whatever map we choose strongly biases the things we look for and don’t look for in a given territory. That wouldn’t be a problem given the capacity to cross-reference other maps, but it’s also true that some people confuse the one and only map they have at their disposal with the territory, and utterly fail to consider anything that falls outside its grid. We’ll have a better lay of the land if we keep a number of maps handy, use the mechanical ones for machines, the organic ones for the organic, and avoid confusing a map of one logical type with a system that works on a logical type of a higher order. The operational status of your Ford F-150 is not a reliable indicator of the health of the Ford Motor Company, which is not an indication of the health of the city of Detroit.
