Joho the Blog
An Entry from the Archives

Peter Galison is a university professor of physics at Harvard. He's giving a Tuesday lunchtime talk. [As always, I'm paraphrasing, getting things wrong, etc.]

Positivists tried to ground knowledge in an accumulation of observations, with a minimum of theory, Prof. Galison says. Science would come in the form of little bricks. The result would be "out of the reach of metaphysical theories." Observation-based science would get better over time.

After WWII, via Thomas Kuhn and others, there was a rebellion against the positivist view. Theory comes first, they said. Science was so framed by theory that what counted as valid observation was dictated by the framework of theory. There is no neutral observation and there's no raw perception outside of the framing provided by our theories. Various theories therefore were not continuous (as for the positivists) but were ships passing in the night...at least according to this point of view.

Example: The positivists saw special relativity as the capstone of a continuum of observation-based theories, while the Kuhnians think Einstein overthrew his predecessors and created a new whole.

Prof. Galison looks at the rhythms of the rise of theories. There are breaks in the strands of experiment, theory and instrument but the breaks don't occur at the same time. And that's to be expected because new instrumentation takes a while to yield new experiments and theories.

Doesn't this just make the Kuhnian predicament worse? Now there are three strands with discontinuities, not just the strand of theort. "How do subcultures of science coordinate? What is shared between experimentalists and theorists, or between instrument makes and theorests...or between a subculture like instrument making and the wider technical world?" When a string theorist want to talk to a biochemist, do they have to engage in radical translation as posited by the anti-positivists? No, it's more like the pidgins, jargons and creoles that build up at the real, "thick boundaries" between the natural languages. Prof. Galison wants to use linguistic anthropology to see how scientific disciplines talk. He calls the areas where these inter-languages are built up "trading zones."

(There are no pure fields, he says. Physics, for example, contains elements of craft, math, experimentation, Plato...)

He looks at the growth and change of language, its local connections to people and places, and its contextuality within the wider world. E.g., Einstein had the idea that time is nothing but properly coordinated clocks. Poincare in 1898 had to figure out how to synchronize clocks so he could figure out the longitude of places around the world. He used this to talk philosophically about what simultaneity means. In 1900, he saw that simultaneity existed in the intersection of philosophy of time, technology of time synchronization, and the electrodynamics of moving bodies. New trading zones are emerging: Nanotech (surface chem, elec eng, atomic phyics...), string theory (geometers, field theories), simulations (computer science, stat, viz display). Each of these have had to develop jargon, pidgin and then a creole.

Of course these fields are collaborative. The question is how they're able to. We should dig in to understand the shared techniques, theoretical notions, and instrumentation, and how they relate in the various intersecting fields. What exactly is the coordinative project, how does it change over time, and what does that tell us about the local knowledges?

Q: (ethanz) To what extent have people looked at languages between the engineering and business communities?
A: There's been a little work. It's a very important area for research. You can see the influence even in the architecture. When a lab is built, it reflects where ideas sit in the world. Does the lab resemble a church? A factory? Is it a place to bring VC's? "VC's don't expect to go to PS 101 circa 1956. That's not who they expect you to be."

Q: (eric von hippel): When engineers talk to others, there's often someone translating. What are the general principles?
A: There may not be a generic solution to this problem. In the coordinative process, what is characteristically coming from the different groups. E.g., in the development of radar, the engineers taught the physicists to look at the problem as that of providing a black box with voltage in and voltage out. So, look at what's relevant to the exchange — part of it is knowing what to ignore.

A: (halley suitt) If two disciplines like biology and chemistry create a whole new set of words to describe biochemistry, per your example, is the number of new words in a new field -- for instance all the Net jargon bursting forth -- predictive of the importance of that new field?
A: Boundaries are always thick. "In these thick regions of exchange, a lot of the vocabularies produced by these individual fields get thrown out. There's a stripping away."

Q: (judith donath) In a consulting company, there's a different type of cooperation — it's very competitive. They're trying to use a new language to convince the clients that they have something unique to offer. It's more about jockeying for leadership than cooperating. It's intellectual scent-marking
A: Yes, complexification can sometimes be obfuscation. Cooperation and competition are closer together than we often think. Freud once said "Ambivalence is the basic emotion." I'm trying to move away from an intentional account.

Q: (me) Do these intersections simplify the concepts? Do the participants translate back into their own terms? Do the translations change the theoretical understanding of each of the fields?
A: There's a difference between simplification and regularization. We know how to talk differently to different people. The regularization doesn't mean that you're simplifying the ideas. When two famous physicists wrote a book on physics, they wrote a second volume for experimenters. All sorts of intra-theoretical connections were dropped out, but the calculations got more complicated. This was more regular but also more complex. So, is it simplification? Simple is not a simple concept. The "outspeak" becomes regularized. And does it feed back into the constitutive fields? Yes. And that's why interests me most. It's not that physics founds everything. The radar story is an example where the engineers taught the physicists. The fact of the coordinative activity begins to backform the science. I think that happens all the time, and it's really important. The top down picture not only is bad business, but it keeps you from learning things.

Q: (wendy seltzer) Attorneys are called on to speak many disciplinary languages. What's the role of formal education in differrent disciplines as opposed to learning on the job? Should law firms hire biochemists to talk to biochemists. Others hire generally interested people. I think the generally interested people are going to better at explaining it to a judge, etc.
A: People often think that's what wanted in a book about science is to make it stupider. That seems to me to be empirically wrong. Pedagogically, we need to develop ways to speak about technical-scientific ideas that are not weighed down by jargon and don't require prior knowledge, but are cognizant of the ability of people to deal with complex concepts. If you go to an art museum, it's designed for adults. If you go to a science museum, it's designed for children. You don't have to write in one-syllable words and assume the audience wants to climb on the exhibit. [You go, Dr. Pete!]