Mathematical research has a dirty secret: we’re terrible at sharing knowledge of failed approaches, and we’re not always honest about crediting the work that enables our breakthroughs. These problems are connected, and they’re costing us enormous intellectual progress.
The Failure Gap
When mathematicians publish papers, they share what worked. What they don’t share, what they can’t share under the current system, is the rich landscape of failed attempts that led to the eventual solution. Every breakthrough sits atop a hidden mountain of dead ends, false starts, and approaches that almost worked.
This creates massive inefficiency. Researchers waste time rediscovering the same failures. More importantly, those failures often contain valuable insights about why certain approaches don’t work, insights that could guide future research or reveal unexpected connections.
But why don’t we share failures? The answer is simple: research is frequently about “credit” and being “right.” Share your failures and you might be criticized. You’re unlikely to get credit for progress on a problem, even if your failure helps the next person solve it completely. To the victor go the spoils.
The Credit Attribution Problem
Even when mathematicians do build on previous work, the credit system encourages a kind of selective amnesia. Researchers have strong incentives to present their work as more original than it actually is. This isn’t necessarily malicious, it’s often a survival strategy in a competitive field where careers depend on appearing to make fundamental contributions.
The problem compounds across generations. Advisors pass these behaviors to students, who learn not just mathematics but also professional norms around credit and attribution. If your mentor doesn’t model generous citation practices, why would you?
This creates what we might call “mathematical amnesia,” important ideas get “rediscovered” repeatedly because their actual developmental history gets obscured. We lose track of the real intellectual genealogy of our field.
The Dream: A Mathematical Blockchain
Imagine a system where every mathematical insight, including failed approaches, gets recorded with immutable attribution. Every time someone builds on previous work, the contribution gets tracked and credited. Failed approaches that later prove valuable get their due recognition. The full collaborative reality of mathematical progress becomes visible.
This isn’t just fantasy. The technical infrastructure exists. We could build systems that track intellectual contributions the way blockchain tracks financial transactions. Imagine research where:
- Every failed approach gets documented and credited
- Partial progress receives recognition proportional to its ultimate value
- The tools and techniques that enable breakthroughs get full attribution
- Cross-field connections become searchable and discoverable
Why It Won’t Happen
The problem isn’t technical, it’s cultural and economic. The current credit system is baked into the fundamental structures of academic mathematics: tenure decisions, grant funding, award recognition, and career advancement all depend on appearing to be the primary driver of original breakthroughs.
Researchers face a collective action problem. Everyone would benefit from shared knowledge of failed approaches and more accurate attribution, but individuals have strong incentives not to participate. Why document your failures when competitors might use that information to beat you to a solution? Why share credit when your career depends on individual recognition?
The financial ramifications are real. A mathematician who generously shares partial progress and credits collaborators might find themselves at a disadvantage compared to someone who presents their work as more independently revolutionary.
What We Can Do Instead
Since systematic change seems unlikely, what can individual researchers do within the flawed system we have?
Try to do better at attribution, while accepting imperfection. Most of us genuinely want to credit previous work appropriately, but we’re imperfect at it. The goal isn’t perfection, it’s making honest efforts to recognize intellectual debts while acknowledging we’ll sometimes fall short.
Document your own failures. Even if there’s no systematic way to share failed approaches, keeping personal records of what doesn’t work can be valuable for your own future research and might occasionally help colleagues working on similar problems.
Err on the side of generous citation. When in doubt about whether to cite something, cite it. The marginal cost of an extra citation is low, and it helps preserve intellectual history.
Recognize the collaborative reality. Even when presenting your work, acknowledge that mathematical progress is fundamentally collaborative and cumulative. Individual breakthroughs emerge from complex webs of previous insights.
Mentor better attribution practices. If you advise students, model generous citation practices. Break the cycle of passing down academic survival strategies that prioritize individual credit over intellectual honesty.
The Incremental Solution
Perhaps the most important thing isn’t finding a revolutionary solution to systemic problems in mathematical attribution. Maybe it’s the incremental work of individual researchers trying to do better within flawed structures.
The cumulative effect of many people “trying to do their best” might not transform the whole system, but it could gradually improve how mathematics actually gets done. Not glamorous, but perhaps necessary.
We can’t build a perfect system for mathematical credit, but we can try to be more honest about the intellectual debts we owe and more generous in recognizing the work that makes our own contributions possible. In a field built on the accumulated insights of centuries of mathematicians, that seems like the least we can do.
Looking Forward
The dream of perfect attribution and systematic knowledge sharing in mathematics may remain just that, a dream. But the conversation itself is valuable. Simply recognizing these problems and thinking about them consciously makes us more likely to navigate them ethically as individual researchers.
Mathematical truth emerges through collaboration, even when our credit systems don’t reflect that reality. The more we can align our practices with that fundamental truth, the better our mathematics will be.