Monitoring, Modeling, and Memory

Dynamics of Data and Knowledge in Scientific Cyberinfrastructures

Archive for December, 2009

Messy science exposed

Posted by archer on December 4, 2009

John Tierney recently wrote in the NY Times about “emailgate” in the climate science community, culminating in an investigation and the stepping aside of the director of the Climate Research Unit. I think it is clear both that this is terrible PR, and that the findings about climate change overall are no more or less scientifically sound than they were before. But there are two observations that I want to highlight from this article, possibly useful for our efforts in studying scientific cyberinfrastructure:

1. The hacked records reveal a “data person” wrestling to resolve data quality. Usually, this difficult work that requires judgment calls and nuanced knowledge is hidden from the public. Is keeping that work transparent best? What happens when it is exposed, especially over such a politicized scientific issue?

2. The other interesting issue is around memory. Requests for information from citizens through FOIA-style legal mechanisms were thwarted (and we end up with illegal information access). Scientists did not want to have their raw data seen and scrutinized for fear of misinterpretation. How important is it to keep the raw data vs. sanitized data? When should that data be made visible, especially to the public, ensuring sufficient contextual and knowledge passed along with it to minimize misinterpretation? Should (even some) scientists emails be considered part of the scientific record, or lab notebook?

If you’re interested, Jeff Masters has a blog post following up on the scientific claims of critics based upon the CRU data.

Posted in News | 1 Comment »

Thévenot and Schmidt & Simone readings

Posted by archer on December 2, 2009

This week we have been reading Thevenot 2009 and Schmidt and Simone 1996. (What a contrast in writing! I found Schmidt and Simone to be much clearer and easier to read.)

Thevenot

At the top of pg. 795, Thévenot mentions that the heavy costs of activities related to standardization “prevent the most competent experts from participation in standardization work.” I think this is an important point that we may see play out in the dynamic between scientists and IT people working on scientific cyberinfrastructure. Do scientists delegate important aspects of standardization work to IT people? How do teams ensure that scientists can input into the process at appropriate times. Similarly, if teams can find ways to reduce metadata friction, they may be more successful at obtaining feedback from busy scientists.

I am annoyed by and uncomfortable with the language of “arbitrariness” about choosing forms – although I welcome attempts to persuade me otherwise. I have two problems with this. First, “arbitrary” implies coin flip to me. Forms may be chosen for more reasons than pure technical efficiency (industrial worth), which Thevenot seems to acknowledge. But I don’t think it’s arbitrary if other aspects of worth, or even decision dynamics outside of those, end up leading to a certain decision regarding form. Second, I think we need to be careful not to lump all types of form (or standards, or metadata) together. Some may have much more arbitrariness, or contention, than others. You can insist that their is some ambiguity in any given form – but I think there is more ambiguity in human skin color than in age. Some standards may easily be agreed upon while others are contentious – and it is helpful to focus our energy in allowing continued debate and flexibility in the recording of data about the challenging categories.

One of Thévenot’s core conclusions is that:

“Substantialist reduction tends to inspire the belief that the good being sought has been made real, that once the correct elements with the right properties have been assembled, ‘good’ need not signify anything more than conformity to the formulation of the standard and and its measurement. This reduction omits the disquieting face of the engagement with its dynamic exigencies of having to adjust to once’s dependencies on the environment, and with the perspective of guaranteeing a particular good.”

This I find helpful. Yes, standards serve as an infrastructure. Most of the time we can’t bother thinking through the details of how we should categorize or structure our work, so we just lean upon the pre-decided standard. But in so doing we miss opportunities to reexamine the tensions and perspective that relate to the way the standard was chosen. Thévenot terms this sort of operation as “quietude,” whereas disquietude is the infrastructural inversion that rethinks the standard. He calls this inversion “blinking” – the flitting between the usual quietude and those moments of disrupted disquiet.

In scientific cyberinfrastructure, one of the primary goals is scientific discovery. Opportunities for novel breakthroughs may be particularly suppressed by dominating quietude. It will be interesting to look for ways that the projects we are studying deal with disruption, and when they choose to “blink” and reexamine their standards to make room for discovery.

Schmidt and Simone

I appreciated Schmidt and Simone’s recognition that artifact designers can’t know everything – coordination protocols and articulation work will change – and so artifacts should support both temporary and permanent modification by users of their coordination protocols. (It strikes me that for power-users, open-sourced code affords the ability to both modify it for your own use and contribute more permanent changes/improvements back to the repository.) I also appreciated their perspective of the design object as a sociotechnical one, that we can decide what work to try to address with a technical system and what to leave on social components.

The first 34 pages seem more straightforward; my question is then what to do with the Ariadne notation that they develop and figure 4 (on pg. 190). How can we usefully apply this? It is not obvious to me how to use this on our work, but I would be very interested in discussing this more.

Posted in Current Reading | Leave a Comment »

 
Follow

Get every new post delivered to your Inbox.