Chris Prom: Documenting Science in the Digital Age: What's the Same and What's Different

(host) So it's my pleasure
to introduce our first

Digital Dialogue speaker of the year.

We're very happy to have him.

Chris Prom is Associate Director
of University Archives

and Professor of Library Administration
at the University of Illinois

at Urbana Champaign where he's responsible
for managing digital projects,

supervising archival processing
and overseeing record scheduling.

His research interests address the ways
in which archival users seek information

relevant to their needs
and how they use electronic tools.

Chris was the recipient of a 2003,
2004 NHPRCE Fellowship

and he's also a Fulbright fellow.

He holds a Ph.D. in History
from the University of Illinois

where his dissertation was on Mutual Aid
Societies in late-Victorian Britain.

He's also studied at the University of York
in the United Kingdom.

And for those of you especially
in the library archives world,

it's worth noting that he is
the Publications Editor

for the Society of American Archivists.

So welcome Chris and thank you.

(Chris) Thank you Trevor and thank you all
for being here today.

What I'm going to talk about today
is the outgrowth actually,

the unintended consequence if you will,
of the Fulbright Project

that I had the great opportunity
to be involved with in 2009, 2010.

Which was itself sort of
an unintended consequence,

something I applied for at the last minute
and somehow managed

to have fallen in my lap.

But the Fulbright Program
really is a wonderful opportunity

which I hope some of you
if you ever have an opportunity to do

and get involved with
because it provides you the opportunity

to be exposed to a lot of new ideas
and experiences.

I say it's an unintended consequence
because when I started the program,

the entire basis of it was intended to do
what ended up becoming

more or less this website.

And it was supposed to provide
a set of practical recommendations

that archivists could use to go it alone
and do digital preservation

if you didn't have anybody to help you.

Now it was the unintended consequence
in two respects.

First of all, you probably
won't be surprised to find out

that I discovered you really
can't go it alone.

That there is way too much to be done,
there's way too much to learn,

and there are way too many roles
that need to be covered.

So that was one sense
and the other sense really was

I didn't expect to get interested
in scientific archives.

As Trevor indicated, I was trained
as a Humanities Scholar,

I was trained as a historian.

I did both literary analysis
as part of my work.

I did some quantitative analysis.

I thought what could the potential use
of these scientific archives be

for Humanities research
and humanistic research.

And I ended up discovering quite a bit.

Actually, as part of
the Fulbright experience,

one thing I had the opportunity to do
was to visit the archives

of the Niels Bohr Institute
which is in Copenhagen.

If you visit it someday,
which I hope you get to do,

you'll get the opportunity to meet
the Director of the Archives,

a gentleman by the name of Finn Aaserud
who is trained as a physicist

and also as a historian
and I went to these people

and then Finn asked me
to give a little talk

and he said will you talk a little bit
about scientific archives

and what needs to be done
to preserve under the digital age.

And I said this is fine, I'll come in
and I'll tell you what I think.

And this would be done based on
the results of my research.

And here's what you do,
you try to identify the correspondents,

you go through and do
some curation efforts on it.

You look for the data,
you talk to records creators,

try to find find it.

And then he showed me into Bohr's Office
and he described to me the process

that Bohr used when he was formulating
his research ideas

and he described something to me
that was extremely foreign.

My knowledge of the scientific method
was very shady.

I knew you set up a hypothesis,
you tested the hypothesis,

and I had a very linear
understanding of how

the epistemology of science
if you will, work or how science is done.

And what Finn described to me
was very different than what I expected.

He described Bohr sitting
in the middle of the room,

having people circulate around him,
them tossing ideas back and forth,

testing, conjecturing, arguing, debating,
writing formulas down,

crossing them out,
a continual back and forth

as the fact was trying to be established.

And this immediately occurred to me,
how do you document this?

How could you possibly document it?

And maybe that's even more interesting
than the discoveries themselves.

The scientific process.

I came back from that trip
a little bit confused.

But when I got back to Illinois,
one of my first projects

that confronted me was this.

And you can't really see it
all that clearly,

but essentially this will be a site
that's familiar to any archivist

when you call in after
faculty members ask somebody.

You'll see an office that's left
more or less in a state of disarray.

And we were talking earlier
about the great fortune

to be shown the BitCurator Project
and you can't really see it,

but hiding down,
I guess it's appropriately enough

hiding down in the shadows here,
is about two, three hundred discs,

of various ages, antiquities,
discarded technology formats.

There's one in there that was
a precursor to the CD-ROM image phase.

And what do you do with all of this?

So this is the reality that confronted me.

But looking at this as well,
it really raised an issue for me

that every time the archivist
enters an office or a laboratory

or something like that,
there were a set of activities

that took place there.

It took place within
a particular social environment.

And how do you document that?

About the same time
as all this was going on,

I got interested in trying to understand
a little bit better the relationship

between both the macrosocial environment
in which scientific research takes places

and the microsocial environment.

So let's take a look first
at the macrosocial environment.

I don't think it will be
a surprise to most of you

to know that the place of science
is contested in society today.

The government puts
a lot of funding into science

and the NSF has been putting money
into science for a very long time.

But those budgets are being cut
and there is a tendency to want

to cut those budgets
all for fiscal reasons

but also because,
I think I would be safe to say,

sort of general distrust
of the scientific process.

You can see things illustrating that.

For example, the controversy that arose
over the research practices

at the Climatic Research Unit
at the University of East Anglia,

the "Climategate",

the "Climategate" Controversy
and so forth.

What's interesting to me is
that a lot of people that defend science

assume that the scientific
fact construction process

is very straightforward.

And why don't these people
on the other side of the debate

just accept that the science is
what you say it is.

And I think that's ultimately not
a necessarily productive tack to take.

At the same time, I've begun to feel
that as an archivist, it should be our job

to preserve more about
the microsocial environment

within which scientific research
takes place and to understand

how science is actually done.

So if you look at a picture like this,
staged though it is,

and being one
that I found on the internet.

If you look at it as an outsider,
any outsider would have a set of questions

that they would bring to them initially.

When you look at something like this
or enter a lab like this.

What are all these people doing?

What does this equipment
actually tell them?

How do they know that the results
or the information they get,

the data that they get
from this equipment is accurate?

How is that data interpreted and written
into some kind of a research report?

How is that research
reported at a publication

received within the community?

How is it accepted?

How does it become established?

Which reports, in which publications
are accepted and which in fact become

part of the scientific purpose of facts.

These are types of questions
that we want to document.

That's just by way of introduction,
what I'd like to do

is give about 30 so minutes description
of what we're doing in Illinois

and some areas
that I've become interested in.

Raise a series of questions along the way
and then add a discussion with all of you

about areas in which we can consider
additional work.

So just as an overview
of where I'll be going,

I'd like to give you an overview
of some of the conceptual tools

that we're beginning to use
to capture scientific

and make useful scientific documentation
at the University of Illinois.

I'll also do a little bit of a side path
into the Anthropology

and Epistemology of Scientific knowledge.

Which is an area I've become interested in
essentially because one of my colleagues

pointed me in this direction.

And the usefulness, potential usefulness
of that method for documenting science

and not only science
but the scientific creative process.

Then describe the method we're using
at the University of Illinois

and show specifically
how we're applying it

with a particularly important
scientific research collection

that we're acquiring,
which is the papers

of a microbiologist Carl Woese.

On the methods of scientific documentation
as an archivist obviously I looked

to the archival literature
for my main source of inspiration

and specifically we go back
at the University of Illinois

to a report that was written
by Maynard Brichford in 1969 called

Scientific & Technological Documentation.

Actually I found this to be an extremely
useful guide to getting a sense

of the archival methodology
as it's been applied at my own institution

with research collections
that were acquired before my time.

But it does still have
a lot of usefulness,

in particular in helping archivists
and others to identify,

not only the informational value
of records, but also what they tell us

about the activities of those
who generated them.

Other sources that are really worth
looking at in the archival literature,

I've listed here but let's take
a little bit of a look

at Brichford's Themes because I think,
for those of us, for those of you,

who are not archivists,
it's useful to start with Brichford

because for me, he really does
encapsulate a lot

of what the archival tradition
puts into practice.

I think the first thing to point out
is that like a lot of the early

archival literature this is rooted in
very positivist assumptions

about how archivists can work
and how archivists could work.

Just as an example, on page four,
Brichford notes

that the archivist the responsibility,
assembles the evidence,

makes the diagnosis
and prescribes the remedy.

All without leaving a thumbprint
on Clio's scales.

I think all of us today will regard that
as somewhat of a naive viewpoint.

But regardless there are lot
of really useful things

and in particular one of the points
I've really taken away

from what Brichford wrote
is articulation,

warnings against overspecialization
in archival work.

In particular, knowing that probably
the most difficult

and also important element
of archival work

is getting an understanding
of the research importance

of the person who did the work.

Understanding intellectually
what they're doing.

Why they're doing it
and why it's important.

I think I've had enough interaction
with students now over my ten,

twelve years of time in Illinois
to know there are certain students

that come through and get really good
digital curation data management skills

and train but they really can
miss the boat a lot on why it's important

to preserve this stuff so Brichford
is really good on articulating that.

Brichford, also to my mind,
offers probably the best defense

and articulation of what provenance,
original order and evidential value mean

and why they're so important to preserve.

I'll get to that part in a minute.

The other part that's really useful
for Brichford is he provides

a set of reasonably understandable
evaluation methods that can be used

for a variety of documentation types.

One of the things we'll find when we go in
to a faculty member's office,

whether it's a scientist,

whether it's an English Scholar

whether it's a Historian,

is lots of documentary evidence
in different formats

and Brichford really provides
some very good guidance

for using those, how to work
with subject specialists

to identify value, how to use checklists
and how to use really basic things

that need to be done.

The other thing that was interesting to me
about this book is I did not expect

to find this at all but there's a nice long
section about data management practices

and specifically noting the need
to preserve research data,

a sampling of the research data
and here was really the kicker for me,

actually preserving the programs
that were used to interpret

and parse that data.

Given the fact that this
was written in 1969,

you can't argue with that advice at all.

So it's a really good starting point
and in fact I prefer it in some ways

to some of these other works
that came along later.

The Haas Samuels And Simmons Book
that I wrote came out of NIT.

Again, it's a very useful book.

Useful set of practices,
goes into a lot more detail.

There's a much longer section
on data management practices.

So for those in
the data curation community,

in the e-research community,

I think it's useful to go back
and look at some of the prior work

that has been done in this area.

I would point out one thing
I note about it is,

it's organized around a very idealized
sequence of activities

and so they'll set up this nice neat
sequence of things.

A scholar makes a hypothesis,
a scholar tests the hypothesis,

a scholar revises the hypothesis
or accepts it.

And then it keeps going
through that cycle.

And in fact, that's not
how it works at all.

In fact it doesn't work that way.

And they'll acknowledge that.

But it does color the interpretation
to a certain extent

because if you look
at some of the examples

they provide for instance,

they're just wrong or don't make sense.

So for example,
one of the sample hypotheses is

DNA is structured as a double helix.

That was never a hypothesis actually.

If you read the book I'm going
to point out to you in a few minutes,

you'll find out where that idea
of the double helix came from.

In fact, it came from the complete upset
of a set of assumptions about chemistry,

or because of a side conversation
that Watson had when he was at Berkeley.

Here are some other areas
of archival interest

that I think are worth mentioning

for considering
scientific data preservation.

I would just note a couple of things
as an archivist, I go in assuming

that everything I do is based around
this core archival concept

of preserving evidential value
and those of you

who have studied archives,
this whole thing is new to you

but basically what evidential value
is the quality of records

that speaks to the activities
of the person who produced them.

Not the information they produced
but how they went about producing it,

provides evidence about
how it was produced.

One of the things like I said,
Brichford, I'll just pass this around,

in case someone wants
to take a quick look at it,

notes is that this idea
of evidential value

is really put into practice
using one principle

and two doctrines as he calls it.

The principle being provenance,
that materials are arranged

within an overall classification scheme
that reflects the origins and activities

of a record of whoever created them.

The doctrines being first
"Respect des fonds",

which is maintaining each deposit
as a separate entity,

and the second doctrine being
the sanctity of the original order

that when you receive things
from a records creator,

you shouldn't really reorganize them
because by doing that

you may destroy relationships
between them.

Now I'll return a little bit later
to what I think these doctrines are,

are good doctrines,
they're effective doctrines,

but they're limited in their usefulness
in the current environment.

So I just wanted to set those up
as a baseline.

These ideas, I would say,
when you combine them with

digital preservation literature
and with particularly trying

to implement them in light of things
like the OAIS reference model,

Open Archival Information System
reference model,

it can take you quite far.

I was talking to Matt and Trevor,
people ahead of time

and describing what we were doing
at Illinois before we got

a digital preservation coordinator
and before we had proper

digital preservation
infrastructure in place.

And it was essentially what the results
of my research project

when I was on Fulbright Fellowship
were intended to cover

and it's described in this article
which is basically sort of a

here's a do it yourself guide
to digital preservation

for archivists in the absence
of other information technologies.

Now when I look back at that,
I looked back at it the other day,

and like a lot of things
I've written in my life,

I say, "This is incredibly naive,
how did this thing ever get published?"

And essentially,
I structured it as a problem.

The problem being, like a lot of archives,
we have all this stuff we've taken in,

from scientists, from philosophers,
from whoever, it's on discs,

and we don't really know
what to do with it,

so we put it in boxes
and left the problems,

let tomorrow worry about itself.

So that's the problem.

And, here I've got the solution for you.

I would say that things didn't exactly
turn out to be as simple

as I thought they did.

But in general, the process
can take you so far

and I think most archivists
would do very well,

to sort of study
digital preservation literature.

Essentially what we ended
up doing at Illinois

with some of the initial collections

that we had received were to overwrite,
I guess you could say,

to bring our practices into compliance
with the Open Archival Information

System reference model.

So we tried to be very explicit
without having

a supporting computer infrastructure,
other than a file server

and whatever we could install
on our desktop machines,

to keep all of this information.

Whatever it is we're trying to preserve,
in its more or less original order.

Enough information to know
that this is a certain file type

and should be open using this software.

Here's a little bit of "premise-like"
preservation, description, information,

but it's stored in a spreadsheet,
and we may keep importing it

to the big system, it'll solve
all our problems at some future point.

Here's a descriptive record
in our database,

and they're related to each other
because we have an ID that links

the descriptive information
back to the archival family.

So that's what we ended up
implementing in this.

More or less what I teach.

One of the things wasn't mentioned
was I also teach arranging and description

of electronic records course through
the Society of American Archivists.

And that's more or less
what I teach through that course.

Here are all the concepts,
here's what you can do on your own,

without having
the supporting infrastructure.

So we implemented an infrastructure
like this this using our existing systems.

Using some tools like the [inaudible]
with Data Excessioner,

storing it in the [inaudible]
network storage,

using our archival management system,
storage platform as a set of access tools

to provide access to materials.

And it pretty well worked okay.

To go back to the picture I had here
of this office, the Sam Smith office.

This looks pretty disorganized
and it is very disorganized.

It just doesn't look that way.

My favorite thing, you can't see it
off the side here,

but to reuse slide trays at one point,
they took every single slide

that he had used in his teaching
and emptied it en masse

into the top drawer of the file cabinet.

This is kind of what we're dealing with,
but it worked okay applying

the Brichford method
with these OAIS concepts.

At the end of the day, what we end up with
is something that looks a lot like this.

Stanley Smith papers,
here's a description of them.

We have a traditional finding,
which in this case is a PDF file.

It's not an encoded nicer coded
archival description document.

But we also have some nearline thing,
we also have some electronic records,

we have some AV materials
which were actually

really interesting records,
they were captures from

a very early computer aided
instruction system

that Illinois had developed
back in the 1960s and early 70s

for teaching chemistry.

And we also have both a nearline copy
of all his digital files,

and an online copy of it.

Some of it, the online materials being
that which we could put on

Youtube copyright clearance.

And the nearline being everything else
that was either too big

or we couldn't put it online.

So essentially what we end up with
at that point is something

that's possibly ready
for somebody to go along

and do some nice data archaeology on it.

This is essentially Smith's files
as he kept them on his computer

with a little bit of organization
but not much provided by us.

And people could just go in and browse it,
using a little file browsing application.

I haven't shown this here,
but you can also download all this

as a zip file if somebody
wants to apply data curation

or data mining tools to it
that can be done as well.

But it occurs to me,
that looking at the disorder

we've started with, we've imposed
a particular order on this

that yes it captures Smith's essence
but perhaps doesn't reflect it.

Lately we've been becoming
more and more interested in

taking in concepts
from related disciplines,

in the University archives
that we haven't had

so much exposure to in the past
and seeing how we can do

additional work with this material.

So here are some of
the data curation pieces

that I'm currently looking at reading,

I was saying about
the White House directive,

but I'm always looking for new things
in the data curation area,

because we get the stuff
that hasn't been curated very well

and then the question is,
what sense do you make of it?

And that certainly comes into play
with the Woese materials.

We've also begun investigating
the potential usefulness

of digital humanities tools.

In particular, in helping us
in the archives do the assessment

of the value of the materials
for preservation.

One of the big issues we have
is we're moving

personally identifiable information.

We also need to get a sense
as to what the topical areas

that are covered by materials
and electronic format are.

And those are some where I think,
we can look at some really useful,

useful sorts of tools here.

Also interested in partnering
with people who use materials

to use publishing platforms
and are interested in whether

publishing platforms like Scalar
can be used as part of

our appraisal methodology as well.

And all of this is, I think good,
it's fine, it's going,

it works pretty well in principle.

But we're still missing something
because I don't think

we really have documented
the scientific process in a way

that it needs to be done,
if we really are interested

in preserving evidence about how
science was done in a lab.

I've had a great fortune to have
the College of Engineering at Illinois

fund the project to organize
the records of the college.

And we hired somebody
from the University of Texas,

by the name of Bethanny Anderson,
who's been going about and doing that.

Immediately the college came and said,
well look we have all these faculty papers

that we need something to do with
and we'd like you to take it

off of our hands for us.

An office is ready to be cleaned out
and she began to say, well,

this is good but hasn't even had training
in anthropology and sociology,

she said, you're missing the boat
on an entire set of literature

that can be bought to bear
on this problem.

So she pointed out to me really
a seminal book.

One of the things I get to do at Illinois
when you're promoted to being faculty-

from assistant to associate
or associate to full,

is to select a book which you'd like
to have donated to the library.

And this thing's been weighing on me
for about a year and they think

there's not a book,
I can't think of a book,

I've got to think of a book.

Well, I found a book.

And I'll probably, two years from now,
I'll probably look back at it

and say, well I want
to throw all that outside,

but where I am at right now,
this book is helpful, very helpful.

What it is, is it's called Laboratory Life:
The Construction of Scientific Facts.

It's written by a French sociologist,
who I had subsequently found out

will be keynoting
the Digital Humanities Conference

in Switzerland next summer,
which I'm hoping to go to.

And British, I forget if
he's a sociologist

or what he is, Steve Woolgard.

Latour is an interesting guy
because he got a Fulbright Fellowship

to go over to the Salk Institute
and he managed to talk his way

past Jonas Salk and said look, I'd like
to put you guys under the microscope

for a year and see how
a scientific lab actually works.

He says fine, we'll let you have
unimpeded access.

You can observe our scientist,
we'll turn you over to a lab

that's studying a particular compound
that is instrumental in translating

neural activity to the endocrine system.

So he has to learn this whole area.

How does this research work
and how do scientists,

as he puts it, construct facts
in the process of doing it.

Now I'm going to give you
the summary of this entire thing

but like every summary,
it's a misrepresentation of the overall argument.

So if you get the chance,
go back and read the book.

If you're interested in
knowledge creation, epistemology,

scientific processes, anthropology,
it's really worth reading.

And as I say, the key fact question
of this thing, is how are facts produced,

and should say or constructed
in a specific microsocial context

of a research lab.

And being a French sociologist,
being exposed to Foucault

and all of these other French people,
for Latour, scientific facts are as much

a social construction
as any other type of fact.

Which isn't to say
that he's completely a relativist,

but it's really worth looking at.

So I'll pass that around as well.

This is the interesting thing for me
as an archivist,

the emphasis on this book
is on the interaction between people

and the activities that they pursue
in a particular environment.

Now as an archivist, we talk a lot
about provenance

and where records come from,
but I think too much,

my discipline makes
an artificial distinction

between where the records come from
and the activities that people do

in producing them.

So for example, we have a metadata scheme
called a Coded Archival Context.

Corporate bodies, persons and families.

Which you can use to give all kinds
of wonderful description

about the context of records creation,
but there's nothing there to talk about

the activities that people pursue
in producing those records.

Why are they producing those records?

And I think using concepts
such as those articulated

by Woolgard and Latour,
we can point

to our richer set of contextual
relationships and understandings

which I was interested too,
because I know you're working on

Linked Open Data initiatives here,
and this all plays in together,

because it feeds into
the arming of triple concept

and I won't go on too much more.

So what are Woolgard and Latour seeing?

Basically, scientists undertake a process
of constructing facts.

The idea normally is that nature exists.
It has an independent reality.

They don't question that it has
an independent reality.

What they do say is
that independent reality

can only be understood
within a particular set of activities

that take place and within
a particular microsocial environment.

And there are several elements to that.

There's a material infrastructure
which has to be in place.

You might have a device
like a mass spectrometer.

That only exists because of some prior
scientific process which led

to the discovery of the principle
that is then written into that device

to read radioactive decay.

You know the device is reading.

And that construction process
takes place in this environment

but it essentially consists
of slow craft-work.

This is an interesting concept
for me as well,

because I've been reading,
because I've been going back and reading

some Thorstein Veblen lately.

For those of you who have read Veblen,
Veblen placed a lot of emphasis

on productive activity
and on what he calls

the instinct of workmanship.

And you can really see that at play
in the scientific processes.

But these processes take place
under particular circumstances,

a particular time and place.

And one of the main functions
of these processes is to eliminate noise.

Noise for example would be,
in the case of try to isolate

a particular chemical that's used
as say for example a releasing factor,

transmitting neural impulse
to the endocrine system.

There are all kinds of things
that can interfere.

For example, one scientist
made a huge blunder,

because they forgot to account
for the fact that yeast had been injected

from the environment into their culture.

And the scientist's reputation was ruined
as a result of this,

and they didn't believe anything he said.

So you're combating noise
using an agonistic process.

A continual back and forth,
a continual test

and a continual fighting.

In which researchers invested
build credibility.

Both in an intellectual sense
that leads to them getting grants.

In a material sense,
in terms of their pay,

in terms of all kinds of things.

All of these factors worked together,
and without any of these factors

you could not end up
with the result that we have

which is the reification
of one particular factual statement.

So you take all of this away, they say,
and you don't end up

with the reified fact at the end.

DNA has a double helix structure.

We would never know that
if it were not for all of these factors.

So that's what they mean
by the social construction of facts.

Now I'm going to show
you in a minute,

all these concepts,
for me as an archivist,

changed the way I think
about doing appraisal

and looking at research collections
and understanding them.

And I would say, I think
that this is something,

I say I believe it now, but maybe
I won't in several years,

I think they have applicability
for other types of research as well

in the scientific community.

So that's Latour and Woolgard.

I have just put a couple of quotes up here
but basically what they're doing

is taking an anthropological approach.

These conclusions emerged from
essentially a two year long study

where Woolgard was in the lab,
immersed himself in the lab,

watched scientific process taking place.

Was actually given the task of doing
some of the scientific processes.

Hopelessly muddled them up, resulting
in a loss of about a week's worth of work.

Because he didn't have
the craft skills necessary.

But essentially what he's doing
is trying to explain how science

produces these authoritative documents.

Filters out the noise, comes up
with authoritative documents.

Now, really what I think it's useful for,
for me as an archivist,

is to help as an archivist get beyond
the myth-what I would term

the "mythology of science".

I'll show you some examples
of this in the past.

But for those of you who have worked
with scientific collections or seen this,

it'll be familiar to you.

You'll meet with somebody,
they'll tell you how great

this or that scientist was.

It's sort of like the last vestiges
of the great man theory

of scientific discovery.

Everything at the end gets
kind of conceptualized

as this titanic battle
between different models

or interpretative frameworks.

But once that's all over,
and a set of facts is established,

everybody accepts it as a given.

You can think of this as having an analog
and Thomas Kuhn's paradigm notion

that occasionally there are paradigms
that are upset and the theory of relativity

for example set the Newtonian synthesis,
all this kind of thing.

Now Latour and Woolgard are not using
this term mythology

in any kind of pejorative sense.

They're simply saying
that every discipline has

a controlling set of narratives
within which research can take place

and no progress can be made forward
unless everybody accepts those.

Now every once in a while,
somebody overcomes

and upsets the apple cart
but as a result of all of the prior work

that has been done
in this microsocial context.

It's sort of a very playful book,
they make a variety of statements

that are very, I would say,
designed to push buttons.

One of my favorites is: "A fact is
a statement which is accepted by everyone

and is too expensive to overturn."

An expense here meant
not in a pecuniary sense,

but in sense of the overall credibility
of the fact and the creditability

of those who have an investment
in doing a research based on that fact.

In a case of Watson and Crick,
they were struggling to understand

this DNA structure
and they had been relying,

where they relate a really
interesting anecdote about this

on basis of chemical
relationship of molecules

that was completely inaccurate
and a physicist came along

and told them, look you've got to go back,
these chemistry textbooks are all wrong.

You need to look at it
from a new way.

Watson sits down with some
cardboard models at his desk,

starts playing with it, there comes
the double helix structure.

And then they went about testing it.

So it's very interesting,
this process of fact construction.

Now let me provide you
a little bit of background

about how we're applying this at Illinois.

Just by way of background,
the University of Illinois has an archive

but we are Illinois's land grant
University founded in 1867.

Interesting thing about Illinois,
I didn't know this

until several weeks ago,
when we started to get news articles

about the effects of the sequester
on the local economy,

that Illinois is actually
the largest single

NSF recipient by far
in the country.

This was interesting to me.

The University of Illinois Archives
has been around since 1963,

with a variety of systems in place.

You can go look at the website
if you want to.

Some relevant facts here,
we collect both administrative records

for the University
and also faculty papers.

So unlike a lot of University Archives,
we have specific, in fact,

statutory responsibility
in our general roles

to collect these types
of research collections.

And we really push it to the limit
because nobody else

is going to do it sometimes.

We have about 25 thousand cubic feet
of material under management.

About 5 gigabytes of digital material
under management right now.

And you can browse,
we have a variety of systems

we use to provide access to it.

As far as this anthropological approach
or this social fact

construction approach goes,
we are going to be applying this.

We are in the very early stages
of applying this with papers

of a man by the name of Carl Woese,
who was trained as a physicist.

Worked for General Electric,
recruited to the University of Illinois

as a microbiologist.

From the very beginning Woese
is sort of a classic pure research guy.

Interested in the evolutionary origins
of life on Earth essentially.

So what was the primordial soup like,

to put it in gross terms
as I possibly can.

You can read a lot about Woese.

I think a lot of what you read
will fall into this great man doctrine

or theory about how science works.

I brought along for you,
I'll pass these out,

and I'll just give several of them,
those of you who are interested

in keeping it can keep these,
but this is a copy,

which will give you some background
on his research

from Woese's memorial service.

There's a nice description
of his research in there.

And basically what he is known for
is rewriting the tree of life.

So prior to Woese coming around,
the tree of life was based on morphology.

We have these two types of organisms,
bacteria and everything else.

Cells that have new organisms, organisms
that have a nucleus based cell

and ones that don't have
a nucleus based cell.

Woese came along and said no,
we can't understand biology.

This is sort of the mythmaking sense
by morphology, by physical characteristics

we need to understand it
on its genetic basis,

on the nature of the genome type.

Based on his research,
the tree of life was rewritten.

Took about 20 years for this
to be accepted in the biological community

where are there are now
three main branches:

bacteria, archaea,
which Woese discovered,

these were essentially
methane loving organisms.

Discovered only because
Woese knew somebody

that was pulling these stuff
out of the Grand Prismatic Spring

at Yellowstone National Park.

And eukaryotes which are essentially
what we would think of

as higher forms of life.

Now what we've done
with these Woese materials,

we've done the arrangement and description
to date of the physical collection,

which is about 24 cubic feet of material.

We also, very fortunately received
a snapshot of his laptop computer

and extracted 36 gigabytes of files
from that laptop.

I'll describe what we did
with that in a minute.

Before I get to that, I'd like to show you
how some of these concepts

from Latour and Woolgard can be applied.

You're not going to be able
to see this very easily,

but one of the big portions
of the collection

is Woese's correspondence file.

And part of this correspondence file,
in fact, the vast majority of it,

when you see it, you see this constant
back and forth taking place

between Woese and other scholars.

I think in the digital realm now,
maybe not as much of this goes on.

If it does go on, it goes on over email,
it goes on over Twitter,

it goes on in other social media
types of technology.

But here on the left hand side, you have
Woese writing to Francis Crick in 1967

when he was at the Salk Institute
saying hey I'd like to do this study.

And here is Crick writing back
and saying no, you're all wet,

it needs to be done x, y or z way instead.

So there are a lot of instances of this
back and forth argumentative style

of research taking place.

(from audience) It happens on Twitter too.

(Chris)It happens on Twitter yeah.

I mean the question now for us is,
a lot of this has already been lost,

or even if you keep it, on Twitter,
say it's in the LC Archive,

well it doesn't relate to anything.

Somebody's going to be able to look at it,
and there's a whole set of assumptions.

Where do you go and get those?

That's the question.

But it's here with the Woese collection,
the biggest thing I think I took away

from Woolgard and Latour is in the case
of these major research collections,

this guy, I didn't mention this,
but he won,

essentially what the equivalent
of the Nobel Prize in biology is,

they don't actually
give one out in biology.

Though Crawford Bryce,
he's in the National Academy of Sciences,

won the [inaudible] Medal,
all kinds of things.

In the case of people at this level,
you have to keep everything.

And that for me as an archivist,
that's really hard to admit to.

I mean really hard.

Because I know what
the research limitations are like.

Here's some reasons why.

Latour and Woolgard talk about
the material environment within

which scientific research takes place.

Now I came across this thing,
I had no idea what this is.

So I start talking to people
who have worked with Woolgard,

have worked with Woese and they say,
this is what he's famous for.

He didn't invent this
but he made it better.

It's an electrophoresis chamber.

And then they describe to me
the process that he used

to inscribe the genetic code
using this chamber

and using this chamber to produce
two things, such a very dangerous process,

what he would do,
is he would get these organisms,

some kind of bacteria,
he would culture them,

take all the genetic material out of it,
and Woese's assumption was,

if you have a variety
of organisms mixed together,

so it's best to just get a sample
of genetic material from the entire group,

smear it all over a piece of paper
that was impregnated

with some radioactive substance,
suspend that in this thing,

inside a highly flammable fluid
and then run

an electrical current through it
for about 24 hours or so.

You do that twice.

As part of that the various elements,
guanine, all of these things,

would float off to different levels,
of this paper that was impregnated

in sort of a-what's the word
I'm looking for?

Sliding scale of radioactivity.

Where I'd say start off
with less radioactive on one side,

more radioactive on the other.

It'd float off to various levels,
make spots, which you could then X-ray

and then analyse.

So would I have understood that process
when we discovered this device?

No I wouldn't and I asked him,
well how many people

know how to do this now?

He said maybe three or four worldwide
know how to do this.

This is the microsocial environment
that will disappear

and within 50 years nobody will really
remember how this process worked.

It's already been replaced by other kinds
of genetic sequencing technologies.

So here's Woese inscribing.

He's taking the X-ray, he's writing
the particular genetic code on this.

At this point, all these things are
as potential clues or markers.

He did this work for about
twenty years or so,

before he came up with his discovery.

Here's the results
of his inscription process.

Two pallets of X-rays
that I don't have a clue with.

Each one of these corresponds
to a particular organism,

has the name of an organism written on.

But not necessarily what that organism
is now classified as

because everything has been reclassified
as a result of this guy's work.

So you talk about context,
you talk about activities

as much as you talk about people.

Really good example,
I asked them,

what is the most significant one
of these X-rays?

They pointed me to this one.

They said, okay, this is
a Methanococcus organism.

And this was the one.

This was the one that he discovered,
had his eureka moment on in 1976.

And they sent me this scan.

I said, wait a second,
why is the date in 1981?

It's hard enough for scientists to do it.

Now we actually did find the original.

The one where he discovered this organism.

But that's only the first step
of the process.

Here's the second step,
his logbooks.

When he was absolutely
certain of something,

he wrote in this logbook.

He kept one for each organism.

I can't read this one up here,
but here is the basis, the genetic code,

the amino acid sequences
are representative

of a particular ribosome
in this organism.

Eventually these things were transcribed
at another level and inscribed

at another level into,
and again,

you can't see this very well.

We haven't deframed this yet,
but very fragile,

chains of amino acids written on
very large pieces of documentation.

So does any of this make sense
outside of the social context?

Next level of this, Woolgard and Latour
go into this in detail.

Every lab has an element
of literary production

associated with it, otherwise things
would never get published.

Nothing would become known.

Interestingly enough,
the Institute for Genomic Biology

had the foresight
to photograph his lab thoroughly

before it was dismantled.

And this is the other thing,
as an archivist I've learned,

you're dealing with a scientist,
if you don't get in

before that lab is gone,
you're not going to have much later.

Up to this point,
that's not what we've done.

And I think collections
are not as rich as a result.

One of the things Woolgard and Latour
pointed out is,

they have a process of literary production
that is not essentially dissimilar

from that that happens in anthropology,
sociology, humanities fields.

You have a communication
or authoring device

where a manuscript is put together.

You have a set of documents or data
that are external to the lab,

this could be thought of
as the established set of facts.

So I can't see what this is here.

I know there's a phone book down here.

I remember looking at this,
and there were a bunch of copies

of Science magazine there.

There were some other articles,
these are what we're arguing,

the field that we're arguing in.

And we also have data or documents
internal to the lab.

Over to the left here we have
a set of contacts.

This can be thought of
as a stock of creditability.

Those people that Woese could rely on
and where he needed to go to.

And the scientists sit down,
filters all of this, transcribes it,

and we come up with an argument
in the end of a published document.

Interestingly, in the Woese papers,
and I've seen this time and time again,

particularly in biology
but also in physics

and other fields.

We end up with a reprints file at the end.

The reprints file containing
the document itself,

this is one of Woese's more
famous publications,

in all the supporting
documentation for it.

So over here, I'll reprint requests,
you can see who requested copies of it.

There's some correspondence down here,
about the manuscript.

There's some research results included.

This is where I think things like
the doctrine of original order,

that I mentioned earlier,
really do come into play.

If this is dispersed for some reason,
you've lost all the context for it.

So this is really good, again showing
Woese's process and sort of the process

of literary production
in scientific archives.

Another interesting concept
that I think needs more attention

in our appraisal processes
that Latour and Woolgard bring up

is this idea of credit and credibility.

This is interesting because again,
we wouldn't have discovered this,

but for the fact that Woese had titled
one of his folders "the big K".

So I looked at that,
I said right away,

well that's weird,
what does "the big K" mean?

And it was a facetious reference
to the head of the department,

who he had a running battle with.

Interestingly enough, this was
absolutely fascinating to me.

This is after Woese was well known, 1984.

The head of the department, August 16th.

You can't read this,
but on August 16th, 1984

sends him a letter congratulating him
on the receipt of your recent award

from NSF dealing with
the classification of prokaryotes.

The very next day he writes a letter
saying we're disbanding our program

in genetic "epinevolutionary" genomics.

The very next day.

So this is credit and credibility at work
within his department.

And we only know about it
because he had titled his folder

"the big K", which is Stanley Kappel
who is the head of the lab.

Another interesting thing
from Latour and Woolgard,

is this idea of a set of circumstances
that lead to the production of a fact.

If you look at the memorial service
program I handed out,

you very much see the great man,
sort of thing at work there.

That's common, if you read up
almost any article,

Woese is held up as a great man,
but in his research lab,

there's all kind of evidence
about the individual people

who assisted him in his work.

And they're just written out completely.

So if this isn't documented, that's lost.

Reification, extremely
interesting concept I think.

Here is a document that was discovered
that one of Woese's colleagues

mailed to us, that I think
shows his process at work.

Better than just about anything I've seen,
is Carl Woese on Carl Woese.

He's asked to write up
a biography of himself.

And although Woese is very
well-respected, humble gentle man

in a variety of ways, when he was asked
to describe his own work,

I think you'll agree with me,
you see the reification process

at work here.

Yeah I'm getting there.

I'm just about there.

Because I want to leave
plenty of time for--

And here's just another element of that
but you can see sort of this idea

of the titanic ideas battling it out
on the world stage of his work.

Interestingly, it leaves me
with a question too.

As an archivist, how do we prevent
the reification of his work?

And the work of other scientists?

Here we have a nice,
neat description of his papers.

So we want to facilitate the preservation
of a complete research record.

I'll leave with you with just a couple
of final things here.

I want to leave plenty of time
for discussions.

I really would like to learn a lot
from all of you as well.

For me, it's really going to be difficult
to document this scientific process

and make scientific archives useful
for humanistic research

in the digital realm.

I've already mentioned the difficulty
of pulling together things

that are dispersed
within social media technologies.

But even in the case of things
that we get on one single laptop,

it's hard, it's really hard.

The BitCurator project was demoed to me.

We're going to look at how
we can integrate it into our work,

to help facilitate appraisal decisions.

Basically, we're doing a lot of this
internally right now anyway,

by taking a snapshot of Woese's Mac files.

We also found an embedded copy
of a Sun computer on that disc.

So that raised a set
of interesting questions.

But we're doing things like disc analysis.

So you compare, for instance,

I have this nice set
of photographs of the lab.

And that's all great but absent that
and if I try to represent a disc image,

this is what I'm looking at right now.

It doesn't necessarily make
a whole lot of sense to me

at first glance.

How do I sort through this to get at
the really meaningful stuff.

I've got a couple of indications there.

I can see, well here's his mail.

That's obviously not the area
I want to focus on.

And here's a photo library
and stuff like that.

So it's an indication
but that's really all that it is.

I think one of the things
that's proven to be

really helpful for us--

I came across this in the past week
because of one

of our digital curation students,
came to me and said,

"hey Chris, I heard about this Woese thing,
can you give me a copy of his files?"

Yeah, sign your life away
but go to town on it.

And he ran it through
some topic analysis software

and came up with some pretty
interesting results.

Let me see if I can bring--

that's not going to work.

I'll just show the screenshots
in the interest of time.

But anyway, he ran it through
this topic analysis software

and essentially we end up
with a ranked list of topics.

And it's very interesting
but I would say,

one thing that's
a little bit difficult for,

is when you get to the final document,
you really kind of lost the context of it.

At least using
this particular implementation.

So we can look at an email message
or a manuscript,

but we don't necessarily know
other things that were authored

at the same time
or what it relates to.

In case you're interested in this,

I can send a little bit
more information to anybody.

This was written up by Thomas Padilla,
who's in the Graduate School

of Library and Information Science.

And I don't have it on here,
but essentially he'd run this

through five or six pieces of software,

one of which I think was the [inaudible].

Interestingly for me, I found this
in the topic model analysis of Woese.

So a little bit of
a countervailing narrative,

you could say to this,
reified notion of science

that Woese represented.

So you have Woese's public figure
presenting that.

This is actually what he wrote
to the author of a book

in the history of biology,
written for children.

Joy Hawking has written books-
is his writing

on a book called Magnified Wonders.

And I think this really more accurately
describes how Woese viewed himself

that he did when he was asked to give
a description in a public way.

So what I want to close on,
is essentially my position

as an archivist has changed significantly.

And how I go about doing my work,
as some of the things I've found

with doing these scientific collections.

This is how close the talk
that I opened with at the Bohr Institute

and I think looking back at it,
I would say, like I said,

I had naive assumptions
that technology was the main thing,

that it could be done relatively easily,

and that archivists would have to

do it on their own because nobody else

was going to help us
in most institutions.

That's all wrong now,

the work that MITH
and other places are doing

to [inaudible] that.

And here's where I'm at now.

I would say that archives still
are in a unique position

because we have this emphasis
on information evidence

but that we need to bring in
as much conversation,

as much help as possible

from every other discipline
that do want to do

the same types of things
that we as archivists need to do.

So thank you, really glad to be here.

The other thing I learned
when I was over in Scotland,

the big thing is
it's all about collaboration

at the end, it really truly is.

And I think by struggling
with the literature for a year,

or for 10 months and seeing,
hey there's a certain amount of this

I can understand but I really need
to rely on everybody else.

I just look forward to
a really good discussion

over the next half hour or so.

Over much time we've got here.

I'm not sure what your time is.

(host) You've got at least ten minutes,
then people have to get over to--

(Chris) - Yeah I probably should have left.
(host) - 2 o'clock classes and such.

(host) But thank you very much Chris.

(applause)