Proprietary Knowledge: How Patent Law Remakes the Boundaries of Scientific Inquiry

In 1942, the sociologist Robert Merton proposed that science, at its normative core, is governed by a set of institutional imperatives he called the ethos of science. Among the most fundamental of these was what he termed "communalism"—the principle that scientific knowledge is a collective good, the product of a community of inquirers whose findings must be shared openly if the enterprise is to function. Merton was not naive about the gap between norm and practice. But he regarded communalism as a regulative ideal, a standard against which the actual behavior of scientific institutions could be measured and found wanting.

Eighty years later, that regulative ideal confronts a legal and commercial infrastructure that Merton could not have fully anticipated. The biotechnology patent, the proprietary algorithm, the trade-secret research methodology, the data-licensing agreement that restricts how published findings may be used—these are not peripheral anomalies in the landscape of contemporary science. They are structural features of the way knowledge is produced, circulated, and withheld in the twenty-first century. And they raise a question that is at once historical, philosophical, and urgently practical: when scientific inquiry is conducted under conditions of systematic proprietary constraint, does it remain the same kind of epistemic enterprise that the Mertonian tradition described?

The Legal Architecture of Epistemic Control

To appreciate the depth of the problem, it is necessary to understand how intellectual property law actually operates within research environments. In the United States, the Bayh-Dole Act of 1980 transformed the relationship between federally funded research and patent rights, permitting universities and small businesses to claim ownership of inventions developed with federal dollars. The act was designed to accelerate the commercialization of publicly funded research, and by many economic measures it succeeded. American universities now generate billions of dollars annually in licensing revenues, and the technology transfer office has become a standard feature of the research university's administrative landscape.

But Bayh-Dole also inaugurated a structural tension that has deepened with each passing decade. When a university researcher patents a discovery, she does not merely secure a commercial right; she imposes a legal constraint on how other researchers may engage with that discovery. Downstream investigators who wish to build on patented findings must negotiate licenses, pay royalties, or restructure their research programs to work around the protected intellectual territory. In fields where foundational techniques are heavily patented—molecular biology being the paradigmatic case—this can mean that the basic tools of inquiry are themselves subject to proprietary control.

The implications for what philosophers call the "social epistemology" of science are considerable. If the production of knowledge depends on the free exchange of methods, data, and findings among a community of investigators, then a legal regime that restricts such exchange does not merely slow down science; it alters its epistemic character. Knowledge produced in silos, by researchers who cannot share their methods or access one another's data, is not simply slower to accumulate. It is structurally different from knowledge produced in open, communicative communities.

Trade Secrecy and the Disappearing Method

Patents, paradoxically, require disclosure. To secure a patent, an inventor must describe her invention in sufficient detail that a person skilled in the relevant art could reproduce it. This disclosure requirement has historically been understood as the quid pro quo of the patent bargain: the public grants a temporary monopoly in exchange for knowledge that becomes freely available when the patent expires.

Trade secrecy operates on an entirely different logic. Where patent law demands disclosure in exchange for protection, trade secrecy protects information precisely because it is not disclosed. A company that develops a proprietary research methodology, a novel data-processing algorithm, or a unique experimental protocol can protect that asset indefinitely, provided it takes reasonable steps to maintain its confidentiality. There is no expiration date, no mandatory disclosure, no public record.

In sectors where the pace of innovation is rapid and the competitive stakes are high—pharmaceutical development, artificial intelligence research, materials science—trade secrecy has become an increasingly preferred mode of intellectual property protection. The consequence for scientific epistemology is significant: methodologies that are never disclosed cannot be evaluated, replicated, or critiqued by the broader research community. They produce findings, but findings whose evidentiary basis is, by design, inaccessible to independent scrutiny.

This is not a merely theoretical concern. The recent controversy surrounding several major artificial intelligence laboratories—including disputes about whether large language model training procedures should be subject to public disclosure requirements—illustrates how trade secrecy can place entire research programs beyond the reach of normal scientific accountability. When the method is proprietary, the finding cannot be fully evaluated. And when the finding cannot be fully evaluated, its status as scientific knowledge, in any robust philosophical sense, becomes genuinely uncertain.

Universities, Corporations, and the Blurring of Epistemic Identities

The history of American research universities offers a useful perspective on how dramatically the institutional landscape has shifted. Through most of the twentieth century, the research university operated, at least in its self-understanding, as a site of knowledge production oriented toward public benefit. The scientist employed by a university was understood to be engaged in a fundamentally different enterprise from the scientist employed by a private corporation: the former was producing knowledge for the community of inquiry; the latter was producing knowledge for competitive advantage.

That distinction has not disappeared, but it has become considerably less sharp. The rise of university-industry partnerships, the proliferation of sponsored research agreements, and the normalization of faculty entrepreneurship have created institutional environments in which the same researcher may simultaneously occupy the roles of disinterested inquirer, patent holder, and equity stakeholder in a company commercializing her findings. These are not necessarily incompatible roles, but they create incentive structures that can pull in contradictory directions.

Historians of science have documented numerous cases in which corporate research sponsorship has shaped not merely the topics scientists investigate but the findings they report and the findings they suppress. The suppression of unfavorable pharmaceutical trial data, the selective publication of results favorable to sponsoring industries, the contractual prohibition of certain lines of inquiry as conditions of research funding—these are not exotic pathologies. They are recurring features of a research environment in which knowledge production and commercial interest are structurally intertwined.

Toward a Philosophy of Constrained Inquiry

Philosophers of science have developed sophisticated accounts of how social structures shape knowledge production, from Helen Longino's work on the social dimensions of scientific objectivity to Philip Kitcher's analyses of the division of cognitive labor. But the specific challenge posed by intellectual property regimes has received comparatively little systematic philosophical attention.

That gap is worth addressing. The question of whether science conducted under conditions of proprietary constraint constitutes the same kind of knowledge-making as science conducted in open research traditions is not merely an academic puzzle. It has direct implications for how the United States structures its research enterprise, how federal agencies evaluate the reliability of findings produced under industry sponsorship, and how courts and regulatory bodies assess the epistemic authority of proprietary scientific claims.

Merton's communalist ideal may be unrecoverable in its original form. The commercial and legal infrastructure surrounding contemporary research is too deeply entrenched, and the economic incentives too powerful, to be dismantled by philosophical argument alone. But the ideal retains its value as a diagnostic instrument—a standard against which the actual conditions of knowledge production can be measured, and against which the costs of proprietary constraint can be clearly named. Science conducted in the shadows of trade secrecy and patent thickets is not necessarily bad science. But it is, in a philosophically meaningful sense, a different kind of science. Acknowledging that difference is the necessary first step toward thinking seriously about what it means.