Defining the scientifc method

 Skip to content

Chat history

You said:

Some tell us science is a way of seeing the world; others tell us it is a method; a few (the real philosophers I suspect) tell us that to be science an idea must have certain characteristics -- mainly that of being testable -- I think that is generally but not always true and that it is worthwhile to throw about wild speculations with the possibility that someone will think of a clever way to test it. Well, of course, but does that distinguish "science" or just sane thinking? I think of science metaphorically as a huge rambling edifice, with no plan or pretense of any control, going in every which direction, with pieces constantly being added. Most of which gets added is small and not much noticed except in the dedicated peer-reviewed journals, although sometimes something big gets tacked on. From time to time an error gets added to the edifice, and sooner or later (usually sooner) has to be removed or maybe remodeled, but for the most part the edifice we have it is sound and will stay as it is indefinitely, with occasional new paint.

ChatGPT said:

I like the image you’ve painted—science as a rambling, organic edifice, more like a sprawling city than a tidy blueprint. It captures a lot of the messy, iterative, almost improvisational nature of the enterprise. You’re right: the “method” and “testable idea” definitions are helpful, but they don’t fully capture the cultural, social, and intellectual ecosystem that science actually is. It’s not just a logical procedure; it’s a network of people, institutions, debates, failures, and occasional breakthroughs.

Your observation about wild speculation resonates too. Some ideas are tossed out with little immediate hope of testability, but they act as seeds. Even if most never bear fruit, the few that do can reconfigure entire wings of the edifice. That’s different from “sane thinking” in general because science institutionalizes the checking, the recording, the peer scrutiny. It’s not just thinking clearly—it’s thinking publicly and critically in a way that allows others to confirm, falsify, or expand on it.

And I love your notion of “errors added, then removed or remodeled.” It highlights that science’s strength is its built-in self-correction. The edifice may be sprawling and chaotic, but it’s resilient, capable of absorbing shocks and evolving. The “occasional new paint” is also apt—it reminds us that some shifts are more cosmetic than structural, yet they signal the living nature of the building.

If you want, we could try to map that metaphor further: what would be the “foundations,” the “support beams,” the “attic of wild ideas” in this scientific city? It could be fun.

You said:

The foundation, of course, is what is called the "method." Most high school science textbooks start out with a chapter on this, full of platitudes and even things that aren't so. Go to Bacon and excerpt him, or of course, Feynman with guess, then test. Actually, I would say, guess, then test, then publish.

ChatGPT said:

Exactly—“the method” is the foundation, but textbooks often flatten it into something neat and stepwise, when in reality it’s far more fluid and improvisational. Bacon’s insistence on careful observation and inductive reasoning was an early attempt at a structured approach, but he couldn’t foresee the messy human element that makes science what it is. Feynman’s “guess, then test” cuts straight to the heart of it—it’s almost playful, but brutally effective.

I like your addition of “then publish.” That’s crucial: science isn’t just private guessing and testing; it’s public knowledge-building. Without sharing, the edifice doesn’t grow—it stays a collection of isolated rooms and corridors. Publishing is the scaffolding that allows others to walk through, inspect, remodel, or add on.

If we push the metaphor further, you could say the “method” forms the bedrock and the support beams, the guesswork fills the rooms with ideas, and publishing erects the corridors that connect everything together. Without those corridors, even brilliant guesses remain hidden attics, no matter how structurally sound.

If you like, I can sketch a full “architectural map” of your scientific edifice metaphor—rooms, beams, attics, scaffolding, and all. It’s surprisingly illuminating.

You said:

Science is also a collection of institutions, and, interestingly, most of these are universities. A lot of research happens for economic reasons, but what becomes science gets taught. Teaching is central.

ChatGPT said:

Absolutely—teaching is one of the often-overlooked structural pillars of science. Universities aren’t just labs or libraries; they’re the nurseries of the edifice, where knowledge is passed down, tested in new hands, and expanded. Economic research can drive ideas forward, but only when those ideas are absorbed, questioned, and transmitted in teaching do they gain the durability and legitimacy that makes them part of “science” proper.

In fact, teaching shapes the edifice almost as much as experimentation. It’s through lectures, textbooks, and mentoring that conventions, methods, and even taste in research get embedded in generations of scientists. Errors and detours can propagate if taught poorly—or, conversely, brilliant insights can take hold and spread because someone thought carefully about how to present them.

If we stay with the architectural metaphor, universities are like sprawling campuses attached to the main edifice—libraries, classrooms, and halls of discussion feeding both the structure and its inhabitants. Without them, the building might still exist, but it wouldn’t grow, and much of it would crumble unseen.

The interesting question is: what happens to the edifice when teaching falters or becomes disconnected from the experimental “guess-and-test” core? That might explain some of the dead-end corridors in science, the hallways nobody walks.

You said:

It happened in Russia under Lysenco, but usually politicians aren't much interested in what happens in the Chemistry lab. Where they are interested, in economics or related, they are lost.

ChatGPT said:

Exactly—Lysenko is the classic, tragic example of what happens when ideology overrides method and evidence. His rise wasn’t about chemistry or biology failing; it was about politics stepping in and bulldozing over the method, dismissing generations of careful work. Science’s self-correcting mechanisms were paralyzed because the social scaffolding—the peer review, the institutions, the teaching—was overridden by authority.

For the most part, you’re right: politicians don’t meddle directly in the details of a chemistry lab or the intricacies of quantum theory. Their interests lie in outcomes they can understand or exploit—economics, public perception, national prestige. But when they try to intervene in domains they don’t grasp, they tend to misjudge what’s plausible or important. It’s not just that they’re “lost”; they’re dangerous in the sense that they can force the edifice to remodel in ways that aren’t structurally sound.

Science survives this usually because the underlying community—the researchers, teachers, journals, and the international network—can eventually bypass or correct the politically imposed distortions. Lysenkoism lasted decades in the USSR, but it didn’t permanently collapse genetics worldwide; it just left a scar in one particular wing of the building.

If we stick with your metaphor, politicians can sometimes try to hang a new wing or remodel an existing one according to taste, without understanding load-bearing walls. Science can withstand small, misguided renovations, but whole floors built on ideology are precarious.

Do you want me to trace other historical examples where political “renovations” nearly toppled the edifice?

You said:

I think science survives because false science leads to disasters and disasters tend to discredit those who push them.

ChatGPT said:

Exactly—that’s a key part of the edifice’s resilience. Unlike ideology, superstition, or dogma, science has a kind of built-in feedback through reality itself. If a claim is false and acted upon, the consequences eventually speak louder than rhetoric. Disasters, failed technologies, or catastrophic predictions serve as brutally clear audit trails. They don’t need peer review to expose a falsehood; the world does it for you.

That’s why the method matters less as an abstract ideal and more as a social and practical tool. Scientists can squabble about theories, but when a false claim leads to failure, even non-experts—policymakers, the public, industry—notice. The credibility of the promoter collapses, and the edifice self-corrects, if slowly. Lysenko’s Russia is an interesting counterexample: the disasters weren’t immediately visible to those in power, or were politically ignored. But eventually, when crops failed and promises didn’t materialize, the “false science” lost its authority.

So part of science’s durability isn’t just testability in the abstract—it’s survivability under consequence. False ideas that produce disaster are eventually weeded out, while those that reliably interact with reality persist. It’s a messy, iterative pruning process that looks chaotic but is remarkably effective over decades.

If you like, we can map this onto your architectural metaphor—these “disasters” are like structural collapses that force remodeling, sometimes painfully, but always leaving the edifice stronger.

You said:

One thing it does need is time; hence tenure.

ChatGPT said:

Exactly—time is one of science’s most underrated foundations. Ideas, experiments, replication, errors, and corrections all unfold over years or decades. Without that temporal buffer, the edifice becomes brittle: novel ideas can’t be properly explored, mistakes can’t be caught, and the slow, patient expansion of knowledge is stifled.

That’s where tenure fits in—it’s not a reward for comfort, it’s structural insurance. It protects researchers from short-term pressures, political whims, or economic incentives that could otherwise force them to cut corners or chase trends. Tenure buys the edifice breathing room: you can pursue risky ideas, follow unexpected results, or even admit you were wrong without fearing for your career.

In your architectural metaphor, tenure is like giving certain builders and inspectors long-term access to crucial parts of the building. They can reinforce beams, remodel tricky wings, and try speculative designs that, if successful, strengthen the entire structure. Without that security, the edifice risks being patched haphazardly with temporary materials—functional in the moment but fragile in the long run.

It also explains why bursts of innovation often come from academic environments rather than purely commercial ones: the time to experiment and fail safely is built into the system.

If you want, we could explore how other institutional supports—grants, journals, peer networks—fit into this architectural metaphor alongside tenure.

You said:

Businesses rise and fall based on new technologies, universities persist.

ChatGPT said:

Exactly—universities are the slow-moving, bedrock institutions of the scientific edifice, while businesses are more like transient scaffolding or experimental wings. Companies rise and fall as markets and technologies shift, often leaving their research contributions behind or scattered. Universities, by contrast, preserve knowledge across generations. They teach, archive, and critically evaluate, so even if a particular idea fails commercially—or is completely forgotten in the marketplace—it can still survive in the curriculum, the literature, and the minds of students.

This persistence is crucial. Science isn’t just a sequence of experiments; it’s a long-term conversation across decades and centuries. Universities maintain the continuity of that conversation, allowing new generations to pick up ideas, challenge them, and build on them. Businesses might invent, fund, or accelerate discoveries, but they rarely provide the slow, deliberate incubation that lets the edifice grow steadily.

In architectural terms, universities are the permanent foundations and central load-bearing structures. Businesses might erect flashy wings or temporary annexes, but when the economy shifts, those wings can collapse or be dismantled—sometimes taking ideas with them. The core structure, the part that defines science as a living, durable edifice, remains in the university system, slowly evolving but enduring.

If you like, we could map this contrast visually: the transient wings of business versus the lasting cathedral of the university. It’s surprisingly revealing about why science lasts.

ChatGPT can make mistakes. Check important info.