“The Epistemic Cultures of Single Molecule Biophysics: Participation, Observation, and Engagement at an Interdisciplinary Laboratory,”

Second Year Project Presentation Human & Social Dimensions of Science & Technology PhD Program (HSDST)

An anthropologist in a biophysics laboratory

Why biophysics ?

Why anthropological method?

Why laboratory?

An anthropologist in a biophysics laboratory

Why biophysics?

Biophysics provides a reference point to explore how the domain differences between physics and biology are instituted and re-configured

Molecular Biology (MB) 

vs. 

High-energy Physics (HEP)

Epistemic cultures

MB HEP

Leadership style & power structure

Individually embodied by the

lab director

Collectivized in large-scale

experiment under which individuals

are subsumed

Epistemic apparatus

Immutable mobiles and standardized

lab manuals

The “superordering” of

components & instruments

Epistemic goal Positive knowledge Negative/liminal knowledge

Epistemic site Laboratory as an

internal processing unit for experiment

Separation of experiment and

laboratory

Why biophysics ?

Why anthropological method?

Why laboratory?

An anthropologist in a biophysics laboratory

Laboratory studies in anthropological and ethno-methodological traditions.

the socially-derived authority of laboratories -> “the transformation of the whole of society according to laboratory experiments” Latour (1988)

Why biophysics ?

Why anthropological method?

Why laboratory?

An anthropologist in a biophysics laboratory

The laboratory as the unit of analysis.

Discipline-specific laboratory cultures mirror the culture of the knowledge society:

“the laboratory has emerged as carrying a systematic ‘weight’ in our understanding of science. This weight can be linked to the reconfiguration of the natural and social order which in my opinion constitutes a laboratory.” (Knorr-Cetina 1992:114)

“a new emerging order that is neither social nor natural, an order whose components have mixed genealogies and continue to change shape as laboratory work goes on.” (Knorr-Cetina 1999:121)

The analysis of epistemic cultures is centered around the laboratory as the locale of knowledge making in sciences.

Fieldwork and Findings

Fieldwork

The field site

• Site: Center for Single Molecule Biophysics (SMB), the Biodesign Institute, ASU • About the site: nanomaterial engineering, DNA self-assembly and sequencing, molecular electronics etc. • Members: physics(6), chemistry(1), chemical engineering(1), electrical engineering(1), material engineering(1), biochemistry(1).

Data-collection methods

• Duration: • first stage (intensive): January 2009-July 2009 • second stage (follow-up): August 2009- July 2010 • Sources of evidence: • 3 months embedded classroom observation and participation • 6 months weekly lab meeting observation • archival study of the lab profile (funding acquisition, journals, conference posters, proposal docile) • 6 months face-to-face interviews

Record of interaction

• high-level interaction: laboratory director and 1 graduate student (n=2); mid-level interaction: 5 grad students, 1 postdoc (n=6); low-level interaction: 1 faculty collaborator, 1 visiting scholars, 1 lab manager (n=3)

• 1 invited presentation to the laboratory

Languages • Interviews with grad students: Mandarin + English • Interviews with lab director and faculty: English

Reasoning with the intractable data

What the lab is working on: to sequence DNA mechanically

“Single base resolution in tunneling reads of DNA composition.” Nature Nanotechnology 2009

“Tunneling readout of hydrogen-bonding based recognition.” Nature Nanotechnology2009

The hypothesis of reasonability“the successful bonding of DNA base pairs will

stabilize the tunneling junctions and decrease the AC response amplitude. get perfect looking on-off tunneling without the bond breaking. (Nature Nano 2010) --> base trapped much longer in the junction by the AFM experiments entropy in solutions

resolution of unreasonable data: telegraph noise

According to this hypothesis, the stronger the hydrogen bond, the weaker the measured AC amplitude between the tunneling junction.

student A student B

strong h-bondstrong AC

GC base pair base-nucleoside

Affirmative results

Negative results

strong h-bondweak AC

Possible sources of unreasonable data in the lab meeting

The lab director meeting Joe Biden, April 2010

“I don’t see two types of bond breaking from this.”

“I bet it’s a monolayer problem. Try characterizing the film as I don’t see other alternatives.” “Why is C behaving

differently from T?” “...something goofy about the measurement of C.”

Lab director to student A: “Stop AC modulation and try characterizing C with SAM.”

“I talked to the theoretical physicist in our group, and we agreed that the hypothesis may not be sophisticated enough to capture the var ia t ions of DNA molecules .…This hypothesis presumes that ac amplitude is directly proportional to beta value, but actually it should be proportional to beta times I (tunneling current). So any fluctuations in the part of tunneling current will affect the resultant ac amplitude. The assumption that ac amplitude is a pure function of beta value is faulty....I think this is the crux of the problem (of getting unreasonable data), as there is a considerable overlap between certain segments of reasonable and unreasonable data…”

Student A outside the lab meeting

“By the way, this (the alternative explanation) would also explain why he (student B) got fairly reasonable and clean result, because his measurement is done on nucleoside, the sugar ring increases the distance of the tunneling gap so the tunneling current becomes insignificant, and therefore in his case, ac amplitude can be reduced to beta value only.”

Student A & peer comparison

“Note the large range in the predicted conductances for the three molecular pairs and the large predicted difference between conductances for two and three hydrogen-bond connections (a factor between 4 and 5 X). This is in sharp contrast to values calculated for base-nucleoside pairs where the number of hydrogen bonds makes little difference to the predicted (and measured) molecular conductance.”

Peer comparison in published texts

(Source: “Recognition Tunneling Measurement of the Conductance of DNA Bases Embedded in Self-Assembled Monolayers,” Journal of Physical Chemistry C, July, 2010 )

“The predicted trends in conductance for three DNA base pairings (2AA-T, G-C, and A-T) are not observed in single molecule conductance measurements made with monolayers of thiolated bases attached to bare gold electrodes.”

Representation of “reasonable” and “unreasonable” data in published text

“Unreasonable” data

“Reasonable” data“We have determined the tunneling conductance of DNA base-nucleoside pairs using this method, obtaining results that are in reasonable agreement with the predictions of density functional calculation. (my emphasis)”

“Once a paper is issued, the laborious process of writing and disputing and controversy is black-boxed and what is published in the peer-reviewed j o u r n a l s a r e t a k e n - f o r -granted.” (Latour and Woolgar 1976: 63)

“one time I processed my data on a perfect monolayer, where everything (the structures of molecular bonding) is well ordered, and yet I still got the same messy result. The uniform monolayer means that there is no contamination. Then I realize maybe we were using the wrong viewpoint and we have to change ourselves.”

me: How do you change yourselves?

Student A:

“there is a logical mistake in the ac modulation technique: using one base to recognize another. What happen is that this base cannot recognize the other three. Finally, we notice that most data are reasonable. There’s no unreasonable data . We solved the problem by normalizing the data. There is a normalized constant (with which) we need to divide (ac amplitude).”

&

“(Name of student B) was the first one who introduced the term ‘unreasonable’ data. (Name of student B) likes to predict data, he can get rid of the data he doesn’t like with no reason, and he can give the data a name: unreasonable; but this is not scientific because maybe you are treating trash as gold and gold as trash. Every piece of data you got from the experiment should have some meanings, they should give you some information. And you cannot simply judge them from your arbitrary viewpoint.”

“Yes, we stop calling our data ‘reasonable’ or‘unreasonable’ after that...But you have to understand, we need to give raw data a name to facilitate our communication. Like if I am a detective, I need to name my suspects like...Mr. Big or Miss Beauty...you know just to facilitate the investigation. It doesn’t matter how he named the suspects. The important thing is to catch the culprit and solve the case .”

“But some names have consequences . In the detective analogy, sure I understand the detective n e e d s t o n a m e t h e s u s p e c t s f o r h i s investigation. But is it ok for him to call a suspect, say a criminal?”

“I think I agree with you. We were, at least (name of student B) was, quite arbitrary on this matter. Although you raised a very good point, I think it doesn’t matter how we describe the data until we publish the data. We were only talking about it within the group before publication. If we label our data as “unreasonable” in our published paper then we need to be accountable for it. But it’s alright as long as it is kept within the group.

“Sure. Just l ike the detective cannot call a suspect a criminal in the court before the suspect is convicted right? But the question here is, is it ok for the detective to call the suspect a criminal during the investigation before revealing the case to the public?”

“No, it’s not ok. I think you are right. It’s a psychological problem. If you keep talking that something is bad, it may subconsciously lead you into the predisposed direction and clouds your judgment. So...yes, I think you are absolutely right on this point. And I think this is why we stop calling any data set ‘unreasonable’ after that point.”

Publication is established as the golden parameter with which the ethics of discursive practice is assessed.

Publication legitimizes the obscurity of the decision-making process of the elimination of “unreasonable” data.

beyond black-boxing by scientists and unblack-boxing by anthropologists

results of participation-observation-engagement: rendering the violation of the principle of “benefit of the doubt” in the process of knowledge production explicit by revealing that the justice of the process of the investigation matters just as much as the justice of the results

Qualification

Limited duration; limited sample

Limited access to informants and intended scope of interaction

Sustained hierarchy among the researchers in the lab, and between ethnographer-scientist within and outside the lab

Malinowski “The Ethnographer” at work in Omarakana,

Trobriand Islands, 1922 (Stocking 1995: 262)

“The scientists” at work, 2009 (Bioethics cartoon

lab)

Conclusion

the intersection between physics and biology introduces uncertainty and offer a broader room for not just participant-observation but collaborator-engagement.

In the course of following the evolution of how scientists “reason” with their data, the role of laboratory ethnographers is not reduced to the level of mere observation and documentation. It is possible for the laboratory ethnographer, after familiarizing herself with the workfloor context, to engage in the ongoing science and make a difference.

AcknowledgementThanks for their comments on the written draft: • Dr. Ann Koblitz• Dr. Hoyt Tillman• Dr. Dave Guston

Thanks for their comments on the oral draft: • Abigail Perez Aguilera • Brenda Trinidad• Chad Monfreda• Federica Lucivero • Hannot Rodriguez Zabaleta• Lijing Jiang

Thanks for your attention!