Think Differently. Solve the Impossible.

Listen on:
Apple Podcasts | Spotify

The headline is simple: “Weedkiller fight hits the Supreme Court.”

The story people hear is even simpler: A company failed to warn users → people got sick → lawsuits followed.

That’s a collapsed version of what’s happening. In this episode, I break down the structure underneath the Roundup case—not to argue whether the product is safe - to examine how outcomes are shaped:

• What “safe” means and how it’s defined
• Why labels don’t translate cleanly into real-world behavior
• The gap between instructions and how people  use products
• How responsibility moves from manufacturer → regulator → label → user → environment
• The difference between “probably carcinogenic” and “known to cause cancer”
• Whether warning labels change behavior

This isn’t about weedkiller. It’s about what happens when one person’s assumption becomes another person’s exposure—and how difficult it becomes to assign responsibility once that happens. The legal system will decide liability. The deeper question comes earlier:

What did you assume was safe—and who else did that assumption affect?

Prefer reading? → Read the article 

Listen on:

About the Host

Daniel Stih is an aerospace engineer, software engineer, indoor environmental consultant, and author of 12 books. Through engineering, environmental investigations, and interdisciplinary research, he explores how people define problems, interpret evidence, and make decisions under uncertainty. Learn more in Why I Think This Way.

Transcript

Welcome to the Daniel Stee Podcast. I read the headline, Weed Killer Fight Hits the Supreme Court. The issue is whether the manufacture of the popular weed killer Roundup can be sued.

Tens of thousands of Americans are suing the maker of Roundup, and the Supreme Court is set to hear arguments as to whether the company can be sued for failing to warn of cancer risk from glyphosate, the active ingredient in Roundup.

Side note, in this case, the active ingredient is in fact what is said to cause harm. But there is no inverse label for non-harmful, non-active ingredients.

Such are labeled as inert ingredients, classification that is misleading because inert does not mean chemically inert and not harmful. It means inert ingredients can actually be more harmful than active ones.

If you're reading the label, the confusion comes from these terms. They're in regard to product efficacy. That means how well they actually work, in this case, kill the weeds, not human health.

So the headline frames it as company failed to warn people, people got sick, got cancer, lawsuits follow. That's already a collapsed story. It hides the structure.

What did the manufacturer assume? What did the user assume? What did the neighbor assume?

That was next, had the yard next to the person spraying the pesticides, and what did the system, the regulators, the courts, and so forth, assume? Most importantly, who is responsible when assumptions between all of these differ?

Each person has an assumption. How does it change the responsibility? This event's not what people think it's about.

At the surface, the legal fight's about whether the manufacturer failed to warn users about cancer risk from glyphosate. In other words, if the risk existed, people should have been warned.

I they weren't warned, there's some liability, and if they were harmed, they should get some compensation. It sounds clean, linear, legally solid, but it only works because it collapses everything into a single assumption.

Safety should have been known and communicated by the manufacturer.

This is the part I keep coming back to, though. Somewhere in this story is an assumption. I bought this off the shelf at a major retailer.

In one story, it would be Costco. It could be Walmart. Therefore, I assumed it was safe to use.

The manufacturer should have stepped in, warned me. What does safe mean in that sense? I bought it off the shelf.

It was safe to touch, safe to apply occasionally with protective gear, safe over decades of constant use, safe under all interpretations of emerging science. The harder question sits underneath all these assumptions.

Are we really saying that the user, am I really to believe somebody believed a chemical designed to kill a living organism, a plant, was harmless?

Or did they assume something more specific, that it was safe enough under the conditions they were using it, based on what they were told, what was labeled, what was allowed to be sold? That framing matters, because harmless is easy to dismiss.

Safe enough under assumed conditions is harder, more accurate, because once safe becomes an assumption, instead of a defined condition, it doesn't stay contained in the argument.

It goes outward to how it's applied, where it travels in the wind, who else is affected. Label defines conditions under which the product is considered safe to use, theoretically. How many people read labels anyhow?

There isn't a clean statistic like X percent of people read pesticide, herbicide product labels.

The fact is, many of us skim, then read the label fully or at all, and the instructions that are often followed are partially followed, or the instructions are on the label are simplified. And that's not unique to weed killers.

It shows up with medications and power tools and instructions and labels like that. For example, I often see people broadly spraying weed killers, even in the wind.

And the label actually says, don't spray a blanket area, emphasizes spray too wet, not to the point of runoff, and spray directly on the weeds, not the surrounding soil, and to avoid drift and to contain it, and to apply when the wind is low.

The key idea is you're supposed to treat the plant, because it works by getting to the roots of the plant, not saturate the environment.

The user is responsible if you read the safety data sheet on this product for keeping the chemical where it's intended to be used. And in practice, most people don't operate at this level of precision.

They approximate, they spray broadly, they apply it into conditions that are convenient for when they want to do it, not for what's to control it. It's not because they're careless. They're translating instructions into a real world behavior.

But that translation changes the outcome. Most people don't treat instructions as constraints. They treat them as guidance.

So the real world vision, version of this product isn't the labeled vision that the manufacturer had in mind or version. It's the version shaped by people, how people use it in wind, over large areas, with rough estimates of dosage.

First assumption, it's safe because it's on the shelf. The second assumption, using it roughly the way I think it's okay to use it. So the outcomes are not just a function of the product.

They're a function of these assumptions and the execution. And all of this discussion is collapsing it into one.

If harm depends not just on the product, but how it's used, the responsibility doesn't sit in one place. It moves from the label to the interpretation by the user, to the application, and then to the environment, and that changes the outcome.

So the product being judged in the courtroom may not be the same product used. I'm not arguing the product is safe. The product appears to be capable of harm.

I'm arguing that outcomes are shaped by assumptions and execution, and that most discussions around these topics collapse all these into one. This is what I'm noticing. What happens when people operate under different assumptions about safety?

In reality, some assumed it was safe, and they used it, the Roundup. Some assumed it wasn't. I am personally one of those, and I avoided it.

Some assumed it was safe enough for others, meaning your neighbor possibly that sprays it. And so the person that assumed it was safe enough for everybody sprayed it broadly. Hey, it's good for me.

It's probably good for your yard too, and good for our parks and recreation, right? And some assumed the risk transfers, meaning the wind blows, exposure, environment. Don't spray it in the wind.

This creates a multiple layer of assumptions, not a single point of responsibility, of failure.

The thought experiment this raised for me, I had this thought. If somebody can sue the manufacturer, can I sue a neighbor who sprayed it if my neighbor keeps spraying it and I live next to them for 10 years?

Can I sue the city if I go to that park for 10 years? It's a model question. Where does the responsibility sit in the chain of decisions?

Is it the manufacturer that's supposed to do all this, the regulator, the EPA, the label, the user? At each step, there's an assumption and a decision, and a transfer of the risk.

And the legal system tries to assign responsibility somewhere in that chain.

FYI real quick, when they settle these things, they usually assign percent of responsibility, which when I discovered that, it's because they actually can't, they argue about how to split it up, these lawyers.

It's not clean and simple, the manufacturer. They may assign some percent to the user. The thing is, in the end, the user doesn't pay any of the fines.

The manufacturer is the one hit with the million dollar lawsuit, even though they will assign some percentage of responsibility to everybody down the chain. In the end, you don't read about that part.

You only read about how the manufacturer had to pay out a billion dollars or a couple hundred million dollars in a settlement. So the chain doesn't stop the manufacturer.

This is where reality becomes messy, because responsibility depends on which assumption you treat as the binding one. So Model A would be the manufacturer's responsibility, and users relied on that for safety. Model B would be user's responsibility.

Individuals manage their own risk. C would be a distributed responsibility. Risk moves across the people and the environments.

And D would be, we don't know what the risk is, so let's just avoid exposure period altogether. The neighbor scenarios where the clean story, clean responsibility breaks. If a neighbor sprays, they may believe they're acting safely.

They may rely on labeling and regulation for that assumption. And they may even think they're improving not just their property, but the community if they walk down the street and spray the sidewalk to kill all the weeds on the sidewalk.

The exposure doesn't stay on their property, and you did not consent to that exposure because your assumptions about risk were different. I'm the one that assumes unknown risk. I can't be sure.

I'm going to avoid it. So now the question becomes, does one person's assumption about safety extend into another person's environment?

There's another layer of this upstream of the label, and this is a layer that the news coverage skips. The difference between evidence, classification, and what's actually on a label, and this is where the chain of responsibility gets more complex.

The EPA has not required a cancer warning label for glyphosate. At the same time, a World Health Organization body classified it as probably carcinogenic. Those sound like they should lead to the same conclusion.

They don't, because they're answering different questions. One asks, can this cause cancer under some conditions? It doesn't say, doesn't answer what dose, under what exposure pattern, what is the real world risk?

The other asks, does it pose a meaningful risk when used as directed? Different question. So that creates a tension, because now we have two valid models.

Model 1 is the hazard-based one based on the IARC that decides if something causes cancer. If something can cause cancer, warn people, and Model 2 is risk-based, which is the EPA. If something causes harm under expected use, then regulate or warn.

So what should show up on a label, on this label? Possibly carcinogenic? Is probably carcinogenic enough to require a warning?

That's not settled. It depends on which model you think should dominate. Should it be known to cause cancer?

And if it did, would that change how people use it? People don't follow application instructions. They approximate and they rely on I saw it on the shelf as a signal.

So even if a label said known to cause cancer, some people would stop using it. Some would reduce use. Some would ignore it, and some would assume only at high exposure.

Little bit is okay.

So how the chain looks? There's assumptions at every step here. Assumption one, that science can define risk clearly enough.

Assumption two, regulators can translate that science and that risk into rules. Next assumption, that labels can communicate that to people who buy the product.

And next assumption, that is users will actually follow what was communicated and act accordingly, behave accordingly. And once those assumptions break, then the outcome doesn't follow this model.

That's where the responsibility becomes hard to assign. It's not just about the weed killer. It's about what happens when one person's assumption becomes another person's exposure.

The legal system will decide where the liability sits. The deeper question comes before the lawsuit. Why did someone assume something was safe?

Who else did that assumption affect? Once assumptions move beyond my boundary, it's no longer my decision. It's a shared outcome.

Most don't realize when they're making that trade.