Preserving Data & Creating Digital Collections

How do we preserve digital information?

This week Cohen and Rozenweig begin by addressing the issue of longevity of digital information.    Having never thought about this before, I assumed that there was a standard out there somewhere, but it seems there isn’t.  Hardware changes so rapidly that it’s impossible to keep up, and software is even worse.  After considering the obvious question of what is worthy of preservation, they move on to some things that would not have occurred to me, like documenting your website for posterity and leaving some flexibility in the code (XHML versus XML or HTML).  They leave the reader with some very practical tips: don’t design around specific hardware and software, consider future use as much as current needs, and the basics of backing things up (do it regularly and save more than one copy in more than one place).

Cohen and Rozenweig have lots to say about creating historical collections online, from the reasons for building one in the first place (flexibility, accessibility, collaboration and longevity) to what lends itself to an online collection and what doesn’t.  A site focused on, say, submissions from WWII veterans probably won’t attract much interest, where a site on 1980s Casio digital watches (collected by uber-geeks the world over) definitely will.  They tackle the thorny problem of how to solicit contributions, pointing out the paradox that “to build a collection, you need a collection.”  Ultimately, it’s about building trust, being part of a community of interest and (surprisingly), not worrying too much about “qualitative concerns,” which the authors have found to be a non-issue for the most part.  The authors include a helpful case study using their own September 11 Digital Archive, which follows all of the best practices they suggest.

Having covered the basics, the next stop was the gigantic Library of Congress Digital Collections.  I dove into Chronicling America and Liljenquist Family Collection of Civil War Photographs, which contains images of Confederate and Union soldiers and their wives and children, all available in JPEG and TIFF.  In addition to organizing and displaying all this great stuff, the LOC also includes a set of tips for preserving digital information: identify what’s important, decide what to keep, Export what you want, organize the collection, make copies and manage them in different places.  There’s even a downloadable brochure.  

Last, we looked at Flickr’s collaboration with the LOC called “The Commons.”  It’s filled with a huge variety of images from two collections (“1930s – 40s in Color” and “News in 1910”) that are posted with “no known [copyright] restrictions”.  By this, the LOC is indicating two things: “either there was a copyright on the image and it was not renewed,” or “the image is from a late 19th or early 20th century collection for which there is no evidence of any rights holder.”  They are clear that this “not mean the image is in the public domain, but [it] do[es] indicate that no evidence has been found to show that restrictions apply.”

Awkward Civil War Era Photograph.

The other interesting feature of The Commons is the ability to add tags and comments to images.  While this seems to have great potential for illuminating what might otherwise be lifeless photos, there are some concerns about the proliferation of tags and the extraneous comments that people post.  I can certainly understand how the latter might be a problem, but I looked at the last ten images posted and not only were there no goofy comments, but every single one of them in fact had at least one relevant, helpful post.  Perhaps those particular images just haven’t been up long enough to attract the oddballs.

The Center for History and New Media design Omecka (the database we’re using for Spokane Historical)?  I should have known.

I'd Design a Website, but Google Made Me Stupid.

This week Cohen and Rozenweig drilled down into the importance of website design, insisting that “[p]eople … tend to have strong opinions about the layout, colors, fonts, and other design elements of a website—and of all websites.”  The point is to balance form and function.  It should be attractive, but has to be usable as well.  While there is no established canon as such, the authors offer some helpful conventions for backgrounds, print, images and video.

Here is a shining example from this week’s work.

Here is a horrendous example, taken from (embarrassingly) my alma mater’s website.

The contrast (no pun intended) is clear.  Certainly there are issues of budget and expertise involved; AASU is not Harvard, after all, but AASU’s site violates just about every one of the basics.  No worries.  I know the IT guy there.

Importantly, Cohen and Rozenweig don’t stop at the technical stuff.  After you build the website, then it’s about "Building an Audience."  This is not necessarily intuitive, and they give practical suggestions. They encourage paying attention to blogs, getting involved in online communities, asking for links to others’ websites, and being thoughtful and intentional about facilitating return traffic.  Interestingly, they critically assess the value of the typical metrics (hits and page views, for example), insisting that they don’t tell us as much as we think they do.  Instead, they suggest that requested pages, referrers, browsers and operating systems, and traffic patterns yield more useful information in terms of how people are actually using your website.

Next on the list was a helpful compendium of digital archives where, after digging around awhile, I found the two examples above.  The Harvard page is, incidentally, fascinating.  Click “browse” and take a look at all the great stuff!  I spent a couple of hours immersed in Lyman Beecher’s autobiography and correspondence, something I needed for a project I’m working on.

Is Google changing the way we read?

And then there’s Google.  Cohen’s article “Is Google Good for History” contrasts neatly with Nunberg’s “Google Book Search: A Disaster for Scholars.”  Where Cohen answers the question “yes,” Cohen questions whether we have any “assurances … that Google will do this right?”  Nunberg goes on to point out ubiquitous Google errors, particularly in metadata.  Here Nunberg overlaps with Miller’s “Is Google Making Us Stupid?” in which Miller posits that the Internet is changing the way we read, reprogramming us to scan pages for data rather than read deeply for content and connection.  Both Miller and Nunberg chide Google for treating books as electronic objects containing information.  They don’t approach books the way a historian would.  Does this surprise anyone?  Does anyone approach a book like a historian does?  Most irritating of all was the keening sense of entitlement, which assumes that Google should tailor their process to our needs instead of theirs.  The thing is, Google paid for it, so Google gets to do it their way.  A number of the comments reflect this same irritation.

The two classes on using Google Search more effectively were illuminating.  I knew many of the commands, (define:, Ctrl-f, “-,” and quotes of course), but I did not know “filetype:,” “intitle,” “site:,” or lots of others.  Dan Russell’s solution to a seemingly impossible problem by employing good search techniques on Google products was astounding.  Russell also gives some great suggestions on improving Google Search results in Todesco’s article, “How to Solve Impossible Problems.”

The “Historical” sites (Cleveland, New Orleans, and Spokane) were great.  I dug into Spokane Historical, of course, and found Polly Judd Park, where I take my kids all the time!  It’s a two-minute walk from my house.  Who knew it was “historical”?

DIY or Get the IT Guy?

Are there some things we just shouldn't do ourselves?
(From someecards.com)

A large portion of the reading for this week was devoted to the how’s, why’s, and potential benefits and pitfalls of digitization.  The recurring question was, should we do this ourselves?  Certainly there is something to be said for the modern-era historian knowing something about digital formats, websites, navigation, community-based tools like Twitter or chat forums, and maybe even the basics of programming.  But is all of this Do-It-Yourself(erism) (DIY) actually good for the discipline in the long run?

Long before graduate school ate my life, I worked as an audio engineer.  When I started, everything was analog; it took not months, but years to master the equipment and to acquire the skills to effectively record and mix a record.  About fifteen years ago, digital audio made its debut.  As computers got faster, smaller and cheaper, millions of people, who otherwise would have purchased studio time, began to record and mix their own music.  Now everybody is an expert.

It's harder than it looks.

The results have been mixed, to say the least.  On one hand, there’s a lot of great independent music out there that would never otherwise have had an audience.  Artists can collaborate from thousands of miles away, trading recorded tracks via email or peer-to-peer file sharing.  Compressed file formats developed over the last ten years (Mp3, Mp4) have made music more accessible and easier to store.  Digital music is, by popular acclaim, a paragon of democratization.  But was it good for music?  Other than the mass-market appeal, no it wasn’t.

First, recording and mixing are much harder than they look.  It is much easier to do wrong than right; virtually nothing about it is intuitive.  Good equipment is essential, but because this equipment is expensive, the average user quite naturally opts for the cheaper alternative, almost always to the detriment of the end product.  To make matters worse, as we learned in our reading from this week, compressed file formats add another layer of quality loss.  But isn’t it cheaper?  No.  Corrected for inflation, an Mp4 album on iTunes sells for the same price as a 12” vinyl record did in 1970.

Just because we can scan our own documents, create our own websites, clean up our own audio, produce our own videos doesn't mean that we should.  Yes, there are financial issues.  There is no such thing as a budget for these things, so in many cases it’s either DIY or don’t do them at all.  There are, apparently, ethical issues with respect to whom we outsource.  On the other hand, in-house production may lend itself to open access as it exists (perhaps) outside the for-profit, capitalist universe, but that does not make it immune from prosecution (see the EndNote/Zoterolawsuit).  DIY digitization, of course, lends itself to greater involvement in the digital humanities because the steep learning curve provides lots of opportunities for networking and collaboration.

Is there room for a subset of historians who handle the technical end of things?  Alternately, we might collaborate with IT.  I’ve had success doing that while editing digital video for a Public History project using a program with which I was not familiar.  And if he won’t come willingly, we just throw a sack over his head and kidnap him, Comanche style.  Totally doable.

Objectivist Smobjectivist.

From Wikimedia

There are many and varied advantages to the wide availability of historical materials in digital form, and they should not be discounted.  I am, however, an unabashed skeptic.  While there is clear utility to the digital format, to whom is all of this material addressed?  To what purpose will it be put and how does that translate to an understanding of what has been, how it has changed, and what might be?

The big issue as I see it is the old debate over historical objectivity versus subjectivity.  I am not an objectivist; instead, I am a big believer in informed subjectivity.  Digital history takes a big step away from traditional notions of objectivity as it emphasizes the empowerment of a reader who is not limited to a single journal or even a dozen shelves full of them.  In moments the reader can access content that would have been unavailable to him or her just a decade ago, and the new mediums do not impose restrictions on how that material is used, shared or interpreted.  That is a sword that cuts both ways.  In a culture that still struggles with the implementation (or, in some cases the very conception) of egalitarian equality, wider availability is a welcome corrective.  But there are significant problems.

As the lay reader traverses the Internet, he or she makes few distinctions between The History Channel’s website and that of theLibrary of Congress.  Both are treated as though they were of the same standing and are weighed equally.  There is no real assessment of bias, context or reliability.  What the Tea Party says about eighteenth century American history is as valid as a historical monograph.  Wikipedia is the first place many go in search of a concise – and purportedly faithful – narrative of a historical event even though it is notoriously (even hilariously) inaccurate and poorly supported.  This is not to say that the collaborative aspect is of no value; it certainly is, but it comes at potentially significant cost.  The other side of that, however, is that important historical projects have the benefit of wider dissemination and perhaps a greater scholarly audience than they might have previously.  Much of the input I receive for articles submitted for publication comes electronically via email in just weeks rather than months.  Electronic mediums also encourage a dialogue that is notably absent from a formal letter of rejection (of which I have a growing collection).

Perhaps none of this is of much concern because objectivity is largely illusory.  In the end, the point may simply be that history is what we collectively understand it to be.  I would argue that hazards of subjectivity aside, there is no use in referring to an “objective” historical narrative, because no such account exists.  Even the best sources are not, of themselves, truth; they are signposts, written in the hand of an interested observer, that indicate an interpretation of the truth.  Even the most erudite historian has no choice but to interpret an existing interpretation that was, in turn, influenced by the observer’s perspective (itself a product of cultural factors).

And now, here is a picture of a bunny with a pancake on his head.  Interpret subjectively and liberally.

This is my new blog for a course in Digital History.