In March 1960, the Royal Air Force dropped cats out of a plane over Borneo. They came down in wicker baskets, hung from parachutes.
Figure 1. Operation Cat Drop as it is usually retold, from Sustainability Illustrated (2014). Thumbnail illustration from Heart of the Art (2015).
The cats came from Kuching, the capital of Sarawak. The city’s name is often said to mean “cat.” Kuching had been asked for thirty cats and could find twenty-three. Later retellings grew the number to fourteen thousand. The cats were bound for Bario, a village high in the Kelabit Highlands, where rats were eating the rice. The plane made five test runs before the first basket went out. Hundreds of villagers watched the baskets come down into the rice fields. A month later, the village sent a message back to Kuching: “Thanks for the cats – no more trouble with rats and mice.”
The rats were the last link in a chain that started with a good idea. In the early 1950s, malaria teams sprayed DDT and other insecticides on the walls of the longhouses, the long shared houses where families slept. The spray killed the mosquitoes that carried malaria. It also killed tiny wasps. The wasps laid their eggs inside caterpillars, and that kept the caterpillars from multiplying. The caterpillars lived in the thatch, the dried leaves that made the roofs. They could sense the chemical and stayed away from it. The wasps could not, and they died. With the wasps gone, the caterpillars ate the roofs. The village cats died too. With the cats gone, the rats came in.
The spray went on the walls. The damage came later, through the roofs and then the rice. The villagers suspected the insecticide. It took a team from the World Health Organization to trace the damage back to a wasp small enough to miss.
Cities have their own version of the spray. Houston spent $2.8 billion widening the Katy Freeway to more than twenty lanes, to end its traffic jams. The new lanes did ease the jam at first. Then the wider road drew more drivers. Between 2011 and 2014, the rush-hour drive along it got 30 percent longer in the morning and 55 percent longer in the afternoon. Across American cities, economists have found that driving grows in step with new lanes.
A cheaper house farther out lowers your mortgage, and it works. It also adds an hour to your drive, each way. Someone else takes over the school drop-off. Dinner moves later, then apart. On a Friday, a friend texts about drinks. You read it in traffic, forty minutes from home, and text back no. After a few months, the texts stop coming. The mortgage really did drop.
In Sarawak, too, the spray did what it was meant to do. It was a correct answer to a question about one part. The malaria teams asked how to stop the mosquito. A better question was what else the insecticide would kill. That question splits into three, each worth asking before a fix is made, or before the papers for the cheaper house are signed. What else is connected to this part? What does it keep in check? Who decided what counts, and what did they leave out?
What a System Is
Before a concert, one oboe plays a single note, an A, and the rest of the orchestra tunes to it. If the oboe is a little sharp, the whole orchestra is a little sharp. The three questions ask about parts tied together like that. People have tried to define a system for about ninety years. Here are five of their answers, with the orchestra to test them on.
“A system can be defined as a complex of interacting elements.”
Bertalanffy was a biologist. He wanted one language for a cell, an animal, and a society. His next sentence explains what “interacting” means. Parts interact when each one behaves differently depending on what it is connected to. If they behave the same no matter what, “the elements behave independently,” and there is no system. Take a violinist. Alone, a violinist can play at any speed. In an orchestra, the same violinist listens and keeps pace with everyone else. What the violinist plays depends on who else is playing.
“A system is an interconnected set of elements that is coherently organized in a way that achieves something.”
Meadows split her definition into three things to look for. The elements are the things in the system. The interconnections are how they affect each other. The function, or purpose, is what the whole does. In an orchestra, the elements are the players and their instruments. The interconnections are the score, the conductor, and the players listening to each other. The purpose is music for the people in the seats.

Figure 2. The three things Meadows says to look for in a system: elements, interconnections, and function or purpose.
Meadows also ranked the three. The elements matter least. Replace every player on a football team, and it is still a football team. The interconnections matter more. Change the rules to basketball’s, and it is a new game (p. 16). The purpose matters most. She called it “often the most crucial determinant of the system’s behavior” (p. 17), and it is the hardest to see. To find it, watch what the system does: “Purposes are deduced from behavior, not from rhetoric or stated goals” (p. 14). An orchestra may say its purpose is great music. If it plays the same safe pieces every season because they sell tickets, its behavior says the purpose is selling tickets.
Ackoff reached that sentence in three steps. A system has two or more parts. Each part can change how the whole behaves, but never alone. What one part does depends on at least one other part. This holds for any group of parts too. Take the violins out of an orchestra, and the rest do not play the old symphony minus the violins. Every other part now sounds different.
“Systems thinking is a discipline for seeing wholes. It is a framework for seeing interrelationships rather than things, for seeing patterns of change rather than static ‘snapshots.’”
Senge studied system dynamics, the modeling school Meadows came from, at MIT. His book put it in plain words for managers. He defines a skill. Systems thinking is the habit of asking what else a change touches, and when. A snapshot of an orchestra catches one chord. The music is the pattern of chords over time.
Senge also listed laws for what happens without that habit. The first is “Today’s problems come from yesterday’s ‘solutions.’” Another is “Cause and effect are not closely related in time and space.” Houston’s wider freeway fits both.
“System is defined as part co-implying whole.”
Co-implying means each needs the other. Nothing is a part unless it belongs to a whole. Nothing is a whole unless it has parts. For the Cabreras, seeing systems is one of four moves a mind makes. The other three are telling things apart, relating them, and taking a point of view. Ask it of the violinist. Is the violinist part of the orchestra, the string section, or a quartet that plays weddings on weekends? Each whole makes the violinist a different part.
Where They Meet
All five definitions would call the orchestra a system. They agree on four points, though not every definition makes every point.
-
A system is not a pile. Meadows’ example is sand scattered on a road. Add sand or take some away, and it is still sand on a road (p. 12). Bertalanffy and Ackoff make the point from the other side. In a pile, the parts act independently.
-
A system is held together by its connections. Bertalanffy says “interacting,” Meadows says “interconnected,” and Senge says “interrelationships rather than things.” Ackoff’s parts depend on other parts. The Cabreras relate one thing to another. All five point at the links between the parts.
-
A system changes over time. Meadows writes that systems produce “their own pattern of behavior over time” (p. 2).
-
A system is for something. Meadows’ system “achieves something.” The malaria program was for killing mosquitoes. The longhouse was for a dry, safe place to sleep.
Grapefruit juice and some cholesterol pills show all four points. The pill and the juice are not a pile: together, they do something neither does alone. They are held together by an enzyme in the gut that normally breaks the drug down. Grapefruit juice blocks it. A statistician has a name for this: an interaction effect. Two things interact when the effect of one depends on the level of the other. Here, the pill’s effect depends on how much juice someone drinks.
The change also builds over time. The juice disables the enzyme, and the gut has to make new enzyme to replace it. A day after one glass, a quarter of the effect is still there. In one study, three glasses a day for a week doubled the effect on a blood pressure drug. Too much of the cholesterol drug stays in the body. And judged by what it does, the pair is no longer just a cholesterol treatment. It also raises the risk of liver and muscle damage. This is the first question from the opening. What else is connected to this part? For the cholesterol pill, the answer is a glass of juice.
How a System Moves
Yesterday’s glass of juice is still at work today. A credit card has a delay too: the spending happens now, and the bill comes weeks later. The bill can also push back on the spending that made it. Meadows calls it feedback when a change comes back around to push on itself.
There are two kinds of feedback loop, and the same card shows both. A reinforcing loop feeds itself. Interest adds to the debt, and the bigger debt earns more interest. Meadows calls this kind “amplifying, reinforcing, self-multiplying, snowballing” (p. 30). A balancing loop pulls a system back toward where it was. The monthly statement shows the balance, and the cardholder cuts back. Meadows calls balancing loops “both sources of stability and sources of resistance to change” (p. 30).
The card’s delay is a few weeks. In Sarawak, the rats took years to come. A balancing loop with a delay can overshoot and swing back. Someone who opens a frightening bill cuts spending too hard, then splurges when the next bill looks fine. A statistician would see that swing in the data as a cycle. Months close together look alike. Months half a cycle apart move in opposite directions.
This answers the second question from the opening. What does this part keep in check? The answer is usually a balancing loop. Balancing loops are the easiest part of a system to miss, because when they work, nothing happens. Swap the paper statement for an app nobody opens, and the balancing loop is gone. The interest keeps growing.
Why Ask Before the Fix
In Sarawak, every link the spray broke was a balancing loop. Without the wasps, nothing held the caterpillars back, and the roofs sagged. Without the cats, nothing held the rats back, and they reached the rice. The twenty-three cats in wicker baskets were a balancing loop, flown back in by hand.
The malaria program counted mosquitoes. It did not count wasps, roofs, or cats.
The five definitions disagree on one point. For Bertalanffy, a system is out in the world, whether or not anyone maps it. For the thinkers after him, it is also a map someone draws, and that person decides where it ends. Meadows wrote that “boundaries are of our own making” (p. 99). Sarawak shows both. The wasps were real whether anyone counted them or not. Whether they got counted was up to the people drawing the map.
No program can count every insect in Borneo, so something always gets left off. The people who live with what was left off tend to notice first. In Sarawak, those people were the villagers. That is the third question from the opening. Who decided what counts, and what did they leave out?
None of the three questions says to skip the fix. Houston may still have needed its lanes. The cheaper house may still be the right one. The questions add one step before each fix. For the lanes, the step is asking who will move in along the new road. For the house, it is counting the hour each way in the car, and the Friday drinks it will cost.
A Closing Invitation. Every fix in this piece worked the way the spray did. It solved the problem it was aimed at. Its cost came later, and somewhere else. The three questions work on your own fixes too.
- Notice the sag. Pick the fix that worked and still cost you the most. The cheaper car that kept breaking down. The rule at home that ended one fight and started a quieter one. When the cost came, what did it look or sound like? A knock from the engine on cold mornings. A door shut a little harder. How long did it take you to notice?
- Find the wasp. What did the old way keep in check without anyone noticing? The coworker who caught mistakes before they went out, until the day they left. The weekly call with an old friend that kept small worries small. Can you bring it back this week, the way the cats were flown in?
- Walk to the roof before the next spray. Take the decision in front of you now: a job offer, a new hire, a change to how your team works. Who will live with it a year from now? Did anyone ask them? Ask them before you decide.
Over Bario in 1960, the baskets came down under their parachutes while hundreds of villagers watched. Twenty-three cats were on their way to answer the question about the spray, years late.
Where This Came From
Each of the five definitions comes from one of four waves of systems thinking. The first wave, from the 1950s, wanted to predict and steer. It brought Bertalanffy’s general system theory, Norbert Wiener’s cybernetics, and Jay Forrester’s system dynamics. A first-wave team would have modeled the village: so many cats per rat, so many wasps per caterpillar. The second wave, from the 1970s, began when people inside a system wanted different things from it. A second-wave team would have asked whose village it was. The third wave, from the 1980s, asked the third question: who gets left out. A third-wave team would have started with the families under the thatch. The fourth wave, proposed in 2021 by the Cabreras and Gerald Midgley, is still unsettled.
Figure 3. The four waves of systems thinking, after Cabrera, Cabrera and Midgley (2021).
Intellectual Honesty Note. Four readings are this piece’s own. Two are about Sarawak: the malaria teams asking how to stop the mosquito, and the story as broken balancing loops. The other two are the malaria program and the longhouse as two purposes, and the grapefruit pair as an interaction effect. The cheaper house is a hypothetical. The fourteen thousand cats of later retellings do not appear in the 1960 press. The video in Figure 1 tells the popular version, which dates the cat drop to the 1950s and adds plague; this piece follows the 1960 record. Tom Harrisson’s account, that the cats ate cockroaches killed by DDT, is unconfirmed. Other accounts have the cats licking DDT off their paws. Senge’s page number comes from later citations. The wave dates are approximate, and the 2021 chapter went unchecked.
References
Ackoff, R. L. (2015, November 2). Systems thinking speech by Dr. Russell Ackoff [Video]. YouTube. https://www.youtube.com/watch?v=EbLh7rZ3rhU
Bailey, D. G., Dresser, G., & Arnold, J. M. O. (2013). Grapefruit–medication interactions: Forbidden fruit or avoidable consequences? CMAJ, 185(4), 309–316. https://doi.org/10.1503/cmaj.120951
Bertalanffy, L. von. (1968). General System Theory: Foundations, Development, Applications. George Braziller.
Cabrera, D., & Cabrera, L. (2022). DSRP theory: A primer. Systems, 10(2), 26. https://doi.org/10.3390/systems10020026
Cabrera, D., Cabrera, L., & Midgley, G. (2021). The four waves of systems thinking. In D. Cabrera, L. Cabrera, & G. Midgley (Eds.), The Routledge Handbook for Systems Thinking. Routledge.
Cortright, J. (2015, December 16). Reducing congestion: Katy didn’t. City Observatory. https://cityobservatory.org/reducing-congestion-katy-didnt/
Duranton, G., & Turner, M. A. (2011). The fundamental law of road congestion: Evidence from US cities. American Economic Review, 101(6), 2616–2652.
Emma. (2015, December 4). Operation Cat Drop. Heart of the Art. https://www.heartoftheart.org/?p=1006
Food and Drug Administration. (2021, July 1). Grapefruit juice and some drugs don’t mix [Consumer update]. https://www.fda.gov/consumers/consumer-updates/grapefruit-juice-and-some-drugs-dont-mix
Forrester, J. W. (1961). Industrial Dynamics. MIT Press.
Ho, S. S. (2024, April 1). It’s a cat’s life: Airborne felines fight in war against rats. BiblioAsia. National Library Board, Singapore. https://biblioasia.nlb.gov.sg/history/2024/4/cats-rats-parachute-malaya/
Magnin, A. (2014, May 6). Systems thinking: A cautionary tale (cats in Borneo) [Video]. Sustainability Illustrated. YouTube. https://www.youtube.com/watch?v=17BP9n6g1F0
Meadows, D. H. (2008). Thinking in Systems: A Primer (D. Wright, Ed.). Chelsea Green.
O’Shaughnessy, P. T. (2008). Parachuting cats and crushed eggs: The controversy over the use of DDT to control malaria. American Journal of Public Health, 98(11), 1940–1948.
Senge, P. M. (1990). The Fifth Discipline: The Art and Practice of the Learning Organization. Doubleday.
Wiener, N. (1948). Cybernetics: Or Control and Communication in the Animal and the Machine. MIT Press.
