What It Takes to Challenge Conventional Wisdom in Science

Science is often described as a process of discovery: ask a question, design an experiment, follow the data, and revise what we know. In principle, it sounds beautifully straightforward. In practice, it is much harder.

Some ideas in biology become so familiar that they stop feeling like ideas at all. They become the background. They shape the diagrams in textbooks, the language we use in lectures, the assumptions built into experiments, and the way entire fields decide what is plausible. Once an idea reaches that level of acceptance, challenging it is not simply a matter of producing new data. It means asking a community to reconsider something it may no longer even recognize as an assumption.

In some ways, challenging conventional wisdom is harder than discovering something entirely new. When a phenomenon has not been seen before, the field may be surprised, even skeptical, but there is often room for the discovery to exist. There is no entrenched explanation that must be displaced. But when new data challenge a familiar model, the discovery has to do two things at once: establish what is true and loosen the grip of what everyone thought was already settled. That second task can be the harder one.

My lab has spent many years working at one of these boundaries.

For much of modern biology, ribosomes have been viewed primarily as uniform molecular machines: essential, ancient, abundant, and largely interchangeable factories for making proteins. In this view, the most interesting regulatory decisions occur upstream — at the level of DNA, chromatin, transcription, RNA processing, and signaling pathways — while ribosomes execute the final step of gene expression in a relatively passive way.

This model has been extraordinarily useful. It helped generations of scientists understand the central dogma and provided a framework for studying how genetic information flows from DNA to RNA to protein. But useful models can also become limiting. They can make certain questions seem unnecessary, certain observations seem peripheral, and certain mechanisms seem unlikely before they are even tested.

Our work has been motivated by a different possibility: that the ribosome is not always a passive, uniform machine, but can itself be a regulatory hub. Ribosomes may differ in their composition, associated proteins, RNA features, spatial localization, and ability to translate specific classes of messages. In some contexts, these differences may matter profoundly — shaping development, cellular identity, disease states, and how cells adapt to changing environments.

This idea was not obvious when we began pursuing it. In fact, it ran against a deeply held intuition in the field. The ribosome was supposed to be the constant. The regulatory complexity was supposed to lie elsewhere.

That is one of the hardest things about challenging conventional wisdom: the old model is often treated as neutral, while the new model is treated as speculative. The burden of proof is not evenly distributed. A conventional idea can persist for years with relatively little direct testing because it has become the default. A new idea, by contrast, must be supported again and again, across systems, methods, and contexts, before it is allowed to become thinkable.

This asymmetry also changes how evidence is judged.

Part of the difficulty is that definitions themselves can clash across fields. In enzymology and structural biology, specialization is often expected to mean a discrete molecular species with a defined composition, a purified activity, and a mechanism that can be reconstituted in vitro. That standard has been extraordinarily powerful for understanding catalytic mechanisms and molecular machines. But genomics and developmental biology often define specialization differently: by context, cell type, timing, localization, regulatory output, and selective effects on endogenous targets. In these fields, specialization can emerge as a reproducible bias in function across a population of molecules or cells, even before every atomic detail is known.

These different standards can lead scientists to talk past one another. A genomics experiment may reveal that a ribosome-associated factor selectively affects translation of a class of mRNAs in a specific developmental state, while an enzymologist may ask whether a biochemically purified ribosome species has been shown to possess a unique intrinsic catalytic activity. Both questions are valid, but they are not the same question. If only one definition of specialization is allowed, entire layers of biological regulation become difficult to recognize.

Across biology, epitope tags, fluorescent tags, affinity tags, proximity labels, and engineered alleles have been foundational for discovering how molecular complexes are assembled, localized, and regulated. They have helped define protein complexes, signaling pathways, organelle biology, chromatin regulators, RNA-binding proteins, and countless developmental mechanisms. Mouse genetics has transformed biology for decades by revealing what genes and molecular pathways do in living organisms. Conditional alleles, knock-ins, knockouts, reporters, and lineage-specific perturbations are among the most powerful tools we have for discovering biological function in vivo.

Yet when these same approaches are used to ask whether ribosomes themselves may be heterogeneous or regulatory, they are sometimes treated as uniquely inadequate. Tags that are acceptable for studying almost every other molecular machine are suddenly viewed as insufficient. Mouse genetics that is accepted as a gold standard for revealing developmental mechanisms is suddenly dismissed as unable to teach us about ribosome function. Perturbations that would be considered informative in signaling, transcription, chromatin, or organelle biology are judged by a different standard when applied to the translation machinery.

Of course, every method has limitations. Tags can perturb proteins. Genetic models can have indirect effects. Imaging can be overinterpreted. Biochemistry can lose spatial context. Sequencing can average across cell states. These caveats are real, and they matter. But they are not unique to ribosome biology. The question should not be whether any single method is perfect. No method is. The question is whether convergent evidence from orthogonal approaches — genetics, imaging, proteomics, biochemistry, genomics, and functional perturbation — begins to reveal a biological principle that the old model could not explain.

That is how most fields move forward.

In my experience, paradigm shifts rarely begin with a single definitive experiment. They begin with discomfort. A result that does not quite fit. A pattern that appears repeatedly but is easy to dismiss. A limitation in the tools available to ask the question properly. Over time, these small tensions accumulate. Eventually, the field needs new ways of seeing.

That has been a central motivation for our lab: to build tools that allow us to ask questions that older approaches could not resolve. How are ribosomes organized within cells? Are all ribosomes compositionally identical? Do ribosome-associated proteins confer specificity? Can individual ribosomal proteins influence the translation of distinct groups of mRNAs? Are ribosomes positioned near organelles or developmental signals in ways that shape local protein synthesis? Do changes in ribosome composition matter during embryogenesis, regeneration, cancer, or inherited disease?

These questions require more than a change in interpretation. They require new technologies. They require methods that can capture ribosomes in specific places, in specific cell types, and at specific moments in development or disease. They require ways to perturb mature ribosomes without simply disrupting ribosome biogenesis. They require imaging approaches that can reveal organization at the level of individual ribosomal particles. They require genomics, proteomics, chemistry, genetics, and cell biology to work together.

This is another reason challenging conventional wisdom is difficult: the first tools are never perfect. New methods are often judged against standards established by older questions, even when they are designed to make entirely new questions possible. There is a period when the biology, the technology, and the conceptual framework are all developing at once. That period can be scientifically exciting, but it can also be vulnerable. It invites scrutiny, and it should. But it also requires imagination.

There is also a personal dimension that scientists do not always talk about openly.

It can be lonely to work on an idea before it is widely accepted. It can be exhausting to repeatedly explain why a question matters, why an assumption deserves to be tested, or why a result should not be dismissed simply because it does not fit the existing model. At times, the criticism can feel less like skepticism of an experiment and more like resistance to the possibility itself.

I do not say this because criticism is unwelcome. Quite the opposite: rigorous criticism is essential. It has sharpened our experiments, strengthened our tools, and forced us to be clearer about what we are claiming and what we are not claiming. But there is a difference between skepticism that improves science and skepticism that polices the boundaries of what a field is willing to imagine. Living in that space for many years has been harder than I expected.

And yet, those are also the moments when the purpose of the work becomes clearest. What has kept me going is the data, the creativity of the people in my lab, and the conviction that biology is almost always more interesting than our simplest explanations. Again and again, we have seen that the translation machinery is not merely a passive endpoint of gene expression. It is spatially organized. It is molecularly diverse. It is developmentally regulated. It can be remodeled in disease. It can shape which proteins are made, where they are made, and under what conditions.

Importantly, challenging conventional wisdom is not the same as being contrarian. The goal is not to replace one dogma with another. I do not think the right conclusion is that every ribosome is specialized, or that ribosome heterogeneity explains everything. Biology is rarely that simple. The more interesting and useful question is: when does ribosome regulation matter, where does it matter, and how does it change gene expression in ways that cannot be explained by RNA abundance alone?

That is the kind of question that can move a field forward. It does not erase what came before. It expands it.

The history of biology is full of ideas that initially seemed unlikely because they challenged what everyone “knew.” Genes were once abstract units of inheritance before they became molecular entities. RNA was once viewed largely as a messenger before catalytic and regulatory RNAs transformed our understanding of gene expression. The genome itself was once imagined as a relatively linear instruction manual before we understood the complexity of chromatin, three-dimensional organization, noncoding regulation, and epigenetic memory.

In each case, the field did not abandon the old framework overnight. It widened it.

I see ribosome biology at a similar moment. The canonical ribosome remains one of the most extraordinary machines in biology. Its conserved core is essential to life. But that does not mean ribosomes are functionally identical in every cell, at every location, in every developmental state, or under every physiological condition. Constancy and specialization can coexist. The existence of a conserved machine does not preclude regulatory diversity layered onto that machine.

This distinction is especially important in cancer and disease. Ribosomes with different compositions are often described as “defective,” as though any deviation from a canonical ribosome must represent a broken machine. But in many disease states, especially cancer, cells are under strong selective pressure to grow, survive stress, remodel metabolism, and evade normal constraints. In that context, changes in ribosome composition, ribosome-associated factors, rRNA features, or translational control may not simply reflect dysfunction. They may provide regulatory capacity that supports a particular cell state. This framing changes the question from whether disease ribosomes are merely broken to what regulatory programs they enable that may be beneficial for that cell type — and whether those dependencies can be therapeutically targeted.

For me, this has become one of the most exciting lessons of studying translation: some of the most familiar parts of biology may still be hiding regulatory logic in plain sight.

That is why challenging conventional wisdom matters. Not because new ideas are always right. Many are not. But because the assumptions we fail to question can define the limits of what we are able to discover. They determine which experiments are funded, which observations are believed, which mechanisms are considered plausible, and which young scientists feel encouraged to pursue difficult questions.

I often think about trainees in this context. One of the most important things we can teach the next generation is not simply what is known, but how knowledge changes. We should teach them that skepticism and imagination are not opposites. We should teach them that rigor is not the same as conservatism. We should teach them that the best science often begins when someone notices that the accepted explanation is incomplete.

Challenging conventional wisdom requires data, persistence, humility, and resilience. It requires being willing to be misunderstood for a time. It requires listening carefully to criticism without allowing criticism to extinguish curiosity. It requires knowing the difference between a weak idea and an idea that is merely unfamiliar.

Most of all, it requires staying close to the biology, going where the science takes you.

The data do not care what a field has assumed. Cells do not organize themselves according to our diagrams. Development, disease, regeneration, and evolution often reveal layers of regulation that our simplified models could not anticipate. When those layers appear, our job is not to defend the old picture at all costs. Our job is to ask whether the picture needs to become larger.

That is the spirit that drives our work. We are interested in the hidden regulatory capacity of the translation machinery — in how ribosomes and their associated factors help cells decide not only which proteins to make, but where, when, and under what conditions to make them. We are interested in how this regulation shapes embryonic development, tissue function, disease, and cellular adaptation.

The broader lesson extends beyond ribosomes. Every field has its assumptions. Every generation inherits models that were once revolutionary and later became conventional. The challenge is to honor those models without being confined by them.

Science moves forward when we are willing to ask: what if the thing we thought was constant is actually regulated? What if the machine we thought was passive is making choices? What if the background is part of the signal?

Those questions are difficult. They can be uncomfortable. They can come at a personal cost. But they are also where discovery begins.