About
    • AMMDI is an open-notebook hypertext writing experiment, authored by Mike Travers aka mtraven. It's a work in progress and some parts are more polished than others. Comments welcome! More.
Search
Page Contents
Full

trading zone

08 Sep 2026 - 08 Sep 2026
Open in Logseq
    • A concept from STS (I think) that denotes a sort of conceptual zone where people from different fields communicate and interoperate. I learned this term from Jeff Shrager who wrote about how computationalists and biologists worked with each other (in an earlier era where it was not as commonplace as it is today.
    • Expand this definition, how is it used in general, does it apply to human - AI interoperation
      • Ask Claude
        • Here is an expanded definition, covering the concept's origins, theoretical elaborations, and its applicability to human–AI interoperation.

        • Trading Zones

        • Origin

        • Inspired by anthropological studies of interactions between distant cultures, particularly in the context of trade, Galison (1997) introduced this concept to describe a place where different scientific (sub)cultures connect, despite facing communication challenges due to incommensurable worldviews.
        • More specifically,
        • Peter Galison produced the "trading zone" metaphor in order to explain how physicists from different paradigms went about collaborating with each other, and with engineers, to develop particle detectors and radar.
        • The core insight is captured in Galison's own formulation:
        • "Two groups can agree on rules of exchange even if they ascribe utterly different significance to the objects being exchanged; they may even disagree on the meaning of the exchange process itself."
        • In other words, productive collaboration does not require shared understanding — only locally workable rules of exchange.
        • Galison defined the trading zone precisely as
        • "an arena in which radically different activities could be locally, but not globally, coordinated."

        • The Linguistic Analogy: Pidgins and Creoles

        • A key mechanism within trading zones is the emergence of new shared language.
        • Trading zones are spaces in which scientific subcultures, each with its own language, develop the equivalents of pidgin and creole.
        • A pidgin is a minimal working language cobbled together for transactional purposes; a creole is a richer, more stable language that emerges when a pidgin becomes a native tongue for a community. Applied to interdisciplinary collaboration, this means that trading zones can mature from rough "enough to get by" communication into genuine shared epistemic cultures — or they may stay at the pidgin level indefinitely.
        • Exchanges across disciplinary boundaries can also be carried out with the help of an agent: namely, a person who is familiar enough with the language of two or more cultures to facilitate trade.
        • The ability to converse expertly in more than one discipline is called interactional expertise.
        • A concrete MRI example illustrates the stakes:
        • at one point in the development of MRI, surgeons saw a lesion where an engineer familiar with the device would have recognized an artifact produced by the way the device was being used. It took someone with expertise in both physics and surgery to see how each of the different disciplines viewed the device, and develop procedures for correcting the problem.

        • Collins et al.'s Taxonomy of Trading Zones

        • Harry Collins, Robert Evans, and Michael Gorman extended Galison's concept into a formal taxonomy.
        • The phrase "trading zone" is often used to denote any kind of interdisciplinary partnership in which two or more perspectives are combined and a new, shared language develops. Collins et al. distinguish between different types by asking whether the collaboration is co-operative or coerced, and whether the end-state is a heterogeneous or homogeneous culture. They find that the voluntary development of a new language community — what they call an inter-language trading zone — represents only one of four possible configurations.
        • The two axes are:
        • one dimension is the extent to which power is used to enforce trade — this is the collaboration–coercion axis. The other dimension is the extent to which trade leads to a homogenous new culture — this is the homogeneity–heterogeneity axis.
        • This yields four types:
        • interlanguage, fractioned, subversive, and enforced.
        • Furthermore,
        • different modes of collaboration result in different kinds of trading zone, different kinds of trading zone may be "nested" inside each other, and a single collaboration might move between different kinds of trading zone over time.

        • Related Concepts

        • Trading zones are closely linked to two other STS concepts:
        • Boundary objects:
        • A trading zone can gradually become a new area of expertise, facilitated by interactional expertise and involving negotiations over boundary objects (objects represented in different ways by different participants).
        • Epistemic cultures:
        • The concepts of trading zones, boundary objects, and interactional expertise are particularly fruitful for describing projects in which researchers from massively different epistemic cultures are trying to develop a common language.

        • Breadth of Application

        • The concept has been applied well beyond physics labs:
        • Digital humanities:
        • Digital humanities as a field can be seen as a trading zone and meeting place rather than a strained "big tent."
        • Computer science education:
        • Computer science education requires development of trading zones between experts in the social and learning sciences and computer scientists. Each of these communities uses different methods and speaks a different language, hence the need for a creole and also for interactional experts.
        • Environmental policy:
        • Managing environmental systems like the Everglades also requires the development of trading zones.
        • Science diplomacy: Trading zones have been used to analyze the IPCC–UNFCCC interface, where
        • ethnographic work explored the various ways in which the main findings of the IPCC are made relevant to the UNFCCC, highlighting the significant efforts required to align assessments with needs, and emphasizing the fragile nature of the exchanges between the two organizations.

        • Jeff Shrager's Contribution

        • Your memory of Shrager is accurate and well-sourced.
        • Jeff Shrager (2010) wrote "From Wizards to Trading Zones: Crossing the Chasm of Computers in Scientific Collaboration,"
        • a chapter in the MIT Press volume Trading Zones and Interactional Expertise.
        • Shrager's account focused on his efforts to teach molecular biologists enough computer programming to give them independence from programmers
        • — a practical instantiation of building a trading zone to reduce dependency on specialist intermediaries ("wizards"). His BioBike project was a concrete artifact of this:
        • BioBike was the world's first cloud-based omics computing platform, incorporating a rich network of biological and biochemical knowledge and a high performance reasoning engine, all accessible via a simple through-the-web programmable interface, allowing biologists with little computing background to conduct novel complex computations.

        • Does It Apply to Human–AI Interoperation?

        • The concept maps onto human–AI collaboration in several productive ways:
        • 1. Epistemic incompatibility as the starting condition. A 2026 paper frames human–AI information seeking as a case of fundamental incommensurability:
        • human information seeking behavior and AI agent functionality represent fundamentally incompatible epistemological paradigms, stemming from three dimensions: epistemic orientation (learning vs. pattern reproduction), temporal structure (processual vs. instantaneous), and agentic purpose (uncertainty resolution vs. task execution).
        • This is precisely the condition that trading zones are designed to address.
        • 2. The AI as a non-human party in the zone. The classical trading zone involves two human communities negotiating shared language. With AI, one party is non-human, which complicates the mutual-adaptation story — the AI doesn't learn the human's conceptual framework through social negotiation the way a scientist might pick up a neighboring discipline's vocabulary. The "trade" is structurally asymmetric: humans must adapt their representations, prompts, and outputs to fit what the AI can process, while AI adaptation happens only through training, fine-tuning, or prompt engineering — not in-the-moment social negotiation.
        • **3. New interactional experts ("prompt engineers" and "AI translators").** The trading zone concept predicts the emergence of interactional experts who bridge the gap. The recent rise of prompt engineering, AI red-teaming, and domain-specific AI deployment specialists are precisely this role: