A Framework for Interdisciplinary
Collaboration in Science and Technology
Michael E. Gorman
University of Virginia
March, 2008
All blame reserved-do not reproduce or disseminate
without permission
Michael E. Gorman
Qui ckTime™ and a
TIF F (Uncompressed) decompressor
are needed to see thi s picture.
Conference, Society and Journal
Society, 2007
Conference, Mexico
Sept 2006
Journal 2008
Next: Berlin 2008
Michael E. Gorman
Examples of interdisciplinary
• Herbert Simon and his range of
collaborators, including Newell, Langley
• Convergent technologies to enhance human
performance (NBIC + ethics)
• CERN, Human Genome (Thagard)
• Causal inference in disease ecology
Michael E. Gorman
Collaboration can enhance
creativity by providing a larger
and more divergent pool of ideas,
hypotheses & designs
Provided these ideas can be
communicated across paradigms
Michael E. Gorman
Interdisciplinary collaboration
Disciplinary/expertise depth +
Cross-disciplinary knowledge
Michael E. Gorman
Problem of incommensurability
• Participants from different disciplines or
research communities disagree over what is
a problem worth solving, the proper way to
solve it and what constitutes data
• They operate from different paradigms
– Garcia’s studies of taste and classical
conditioning were rejected by learning journals
Michael E. Gorman
Normal science experts
• Classify problems in ways that point to the
solution (judgment)
• Possess algorithms, heuristics and
procedures, some tacit (skill)
• Have a language they can use to share
information rapidly
• But an expert can be like the drunk looking
under a streetlight for the key…
Michael E. Gorman
‘Tis nobler to be brilliantly
wrong than safely right
• Paradigm is provocatively ambiguous
• Incommensurability is overstated--but
experts and stakeholders from different
communities do often operate as if they
lived in different realities
Michael E. Gorman
Kepler as an example of a
paradigm shift
• Planets orbited on spheres, in perfect circles--in
hindsight, a mental model, but at the time, a reality
for most people
• Kepler had access to Brahe’s data (see Langley et
al.’s work on BACON)
• But Kepler could not get the orbit of Mars to fit a
perfect circle (anomaly)
• Eventually, he discovered that planetary orbits are
elliptical, a view incommensurable with circular
orbit models
Michael E. Gorman
Mental models and paradigms
• MMs can be nested: perfect circle model leads to
lower-order models like specific combinations of
epicycles for individual planets
• Failure of one of these lower-order models can
either trigger its revision, or become an anomaly
for the ‘hard core’ assumptions
• Thought experiments are a form of mental
modeling that can trigger revision (Shephard)
Michael E. Gorman
Bell’s Ear Mental Model
Different ways of inducing current
“Follow the analogy of nature. Make transmitting instrument after the
model of the human ear. Make armature a after the shape of the ossicles.”
Michael E. Gorman
Explains why he and Elisha Gray
were not inventing the same thing
• Bell=mental model of how to translate
mechanical motion into undulating current
• Gray=off-the-shelf solution to transmitting
speech. He though of using the ear, but had
never mounted the ossicles on a device, like
• Bell though of telephone as a breakthrough,
Gray and his backers thought of it as a toy
Michael E. Gorman
Trading Zone as solution to collaboration
despite incommensurablity
• Galison-- scientists and engineers develop an
interlanguage (from jargon to pidgin to creole) to
communicate when designing systems like radar,
particle accelerators
– Golden event vs. statistical paradigms
• Nanocajun
• Lambert--Jet Propulsion Laboratory engineers
refer to their negotiations over where to land a
rover as trades
Michael E. Gorman
Interactional expertise can
complement or substitute for a creole
• AIDs activists who worked with scientists
to rewrite research protocols
• Collins (sociologist of science) who gained
enough knowledge of gravitational wave
physics to pass as a member of the
community on an imitation game
experiment, but could not do research
Michael E. Gorman
Imitation game
• A Turing test for expertise
• Collins & a gravitational wave physicist
answer brief questions
• Members of the gravitational wave
community see if they can tell which is the
• Collins wins
Michael E. Gorman
Sample question
• State if after a burst of gravitational waves
pass by, a bar antenna continues to ring and
mirrors of an interferometer continue to
oscillate from their mean positions?
– Collins’ answers were shorter (impatient
expert) and not as likely to have come out of a
textbook (he had to think them through on the
Michael E. Gorman
Boundary Object
• A technology that can facilitate coordination
across a trading zone, although it is
represented differently by participants:
– Biobike (Shrager), a programming language
that can be evolved both by biologists and
– To create it, Jeff had to acquire dual expertise,
but the users should be able to coordinate
without learning each others’ expertise
Michael E. Gorman
Models and simulations can serve as
boundary objects in trading zones
• Models and simulations can differ in
– programming languages & algorithms
– mathematical techniques
– Levels of resolution
• They can provide a platform for
comparison--or exacerbate
Michael E. Gorman
DOME (David Wallace, MIT) as a
boundary object
• Allows product designers and
environmental consultants to collaborate by
linking their different models
• A kind of creole is established by DOME
that sets requirements for input
• Changes automatically propagate
throughout the system
Michael E. Gorman
Taxonomy of trading zones
Co ll abo rati on
Co e rc ion
Ho m ogeneou s
Co m m unit y:
Inter langu a ge
Con tri bu tory expe rti se
S ha red m en ta l m ode l, l ead ing to co evo luti on
Do mi na nt id eo logy
U ser s c onver ted t o d e signe r’s m en ta l
m ode l
Michael E. Gorman
He terogeneou s
E qu al tra din g zo n e:
C reo le
Inter a cti ona l exp e rti se
Bound a ry ob jec t
E nf orce d
U ser s forced to con for m t o
de si gner ’s m en tal mod el
Incommensurability can be an
Metacognition and (in the case of
value disputes) moral imagination
offer alternatives
Michael E. Gorman
Steps toward metacognition &
moral imagination
1. Awareness that one’s reality is a mental
2. Opens up listening to others’ ‘realities’
(mental models)
3. Leading to alternatives
4. That can be evaluated, using data, models,
improvement in quality of life &c.
Michael E. Gorman
Arizona State University’s decision theater
permits development of a visual creole that
allows multiple stakeholders to form a
trading zone by seeing the impact of their
assumptions on the growth of Phoenix
QuickTime™ and a
TIFF (Uncompressed) decompressor
are needed to see this picture.
Michael E. Gorman
Do virtual environments like Second
Life enhance or inhibit moral
Avatar transforming her environment in Second Life:
The model becomes the world
Michael E. Gorman
Michael E. Gorman
Future research
• Ex Vivo imitation game studies of interactional
expertise need to be supplemented by In vivo and
Sub Species Historiae studies
• Same methods can be used with trading zones
– Including ways of mapping their trajectories
– Better taxonomy, including other types of collaborative
Michael E. Gorman