annet dekker capturing online cultures and storytelling as a method
While this is a very good example of what a large institute can do, in most cases the
Wayback Machine proves to be less reliable.8 As also argued by Web historian Niels
Brügger, an archiving process actively shapes and determines how a website is
archived and hence what kind of reconstruction or analysis is possible.9 Not only do
websites, and their copies, often suffer from temporal or technical inconsistencies,
but as Brügger argues 'the archived website is not an exact copy of the one on the live
Web but a unique version as the result of the archival process'.10 This uniqueness
combined with the arbitrary crawls of the Wayback Machine and possible technical
inconsistencies make it an insufficient tool for art historical analyses, let alone for
the preservation of an artwork. For example, inconsistencies can be part of an
artwork, but they may just be glitches, or a case of bitrot. What one does not see or
know cannot be traced by looking at random crawls.
To overcome a single-page history the Wayback Machine introduced Memento, an
API that allows you to move back in time.11 The application allows users to see the
page around the time it was made rather than the present time. In 2011 the Internet
Archive began to use Memento, which makes it possible to use the Wayback
Machine in an 'interactive' mode. In some cases this generates interesting results.
For example, looking at the changes over time to the website mouchette.org by French
artist Martine Neddam and clicking one of the links via the Wayback Machine on
any random date shows a standardized answer with a link that redirects to a random
other part of the site and not necessarily to the one that it would normally go to
(even in the past). In this case the misdirection is interesting because it has always
been Neddam's desire to make navigating the website as convoluted as possible:
"I wanted to get the viewer lost in a very complex navigation, where the
placement of the links was invisible or unexpected."12
Another example, for which I used Memento's ability to recall history, is the work by
Slovenian artist Igor Stromajer. In 2011 Stromajer began to delete his earlier net art
pieces. After annoucing the project Expunction on Facebook and other social media,
some of the reactions expressed concern: "Igor Can't you do something else
to go through your mid-life crisis 13 I tried to trace his deleted works in the
Wayback Machine with Memento. With a few exceptions, I was redirected back to
his present Expunction project page - even in that past I couldn't revisit earlier
instances of his work. I was stuck in a circular present, no past, no memories. Such
inconsistencies pose huge challenges to researchers who are unfamiliar with the
documentation process, in particular around the nature of absences, redirections
and the limitations of the techniques that are applied. Next to a critical assessment
of the methods and tools that are used, as well as a thorough understanding of the
interfaces in which history is viewed, researchers are advised to collaborate with
Web archivists and Web users to create a socio-technical perspective.14 A cultural
perspective is just as important, in particular in cases where artists and other users
have purposely obfuscated or misused a site's 'standard' functionality. This means
involvement of the primary 'Web users' such as the artist(s), or creator(s), of a site.
In the case of Stromajer, his process of deleting many of his online works is well
documented. All the conversations and the discussions around Expunction can be
traced on Facebook. However, these discussions are not saved on the Wayback
Machine, since Facebook is a closed system and the data of individual users cannot
be cached.15 To avoid the dependency on Facebook, Stromajer took screenshots of
the discussions that are still available on his website. Recently, other tools have been
developed to document social media platforms. One of them is the Webrecorder
created by Rhizome, a non-profit organization based in New York. As Dragan
Espenschied, one of the developers, explains:
"Current digital preservation solutions involve complex, automated processes
that were designed for a web made up of relatively static documents.
Webrecorder, in contrast, can capture social media and other dynamic content,
such as embedded video and complex javascript."16
Indeed, Webrecorder is a good tool to capture social media platforms, it records the
posts, the likes and the comments of other users, and the replay functions as if
you're browsing the live site. But users cannot add anything or make any comments
as one normally could do. For example, one of the recordings is of Amalia Ulman's
project Excellences Perfections (2014). For five months and in almost 200 posts,
Ulman acted out a scripted performance that culminated into an extreme makeover,
which she performed on Instagram and Facebook. Playing on the cosmetics culture
and as a comment on the demands social media makes on users' appearances and
experiences, Ulman convinced many of her followers and (artist) friends that what
she was doing was real. The Webrecorder team documented the entire Instagram
performance, including the Instagram interface, to create a faithful re-performance
of the context in which the photos and comments were embedded.17 At the same
time, Ulman reposted everything to Facebook where the discussion and comments
were more intense. As she says in an interview:
"People got so mad at me for using fiction. That was the main critique: 'It wasn't
the truth? How dare you! You lied to people!"18
hoofdstuk 3
8 For more information about the challenges of Web crawling, especially the impact and interactions of
contextual factors see, for example, Emily Maemura, Nicholas Worby, Ian Milligan and Christoph Becker,
'If These Crawls Could Talk: Studying and Documenting Web Archives Provenance', in: TSpace (University of
Toronto, 2018), https://tspace.library.utoronto.ca/bitstream/1807/82840/1/JASIST_IfTheseCrawls-
Preprint-20180122.pdf, last accessed 13/10/2018.
9 Brügger 'Website History and the Website as an Object of Study', p. 126.
10 Brügger 'The Archived Website and Website Philology: A New Type of Historical Document?', p. 156.
11 For more information see, http://mementoweb.org/about/ and http://timetravel.mementoweb.org/about/,
last accessed 13/10/2018.
12 Annet Dekker, 'In Search of Unexpected', in: Cathy Brickwood and Annet Dekker eds. Navigating e-Culture
(Amsterdam, Virtueel Platform, 2008) pp. 66-68.
144
13 Annick Bureaud https://www.facebook.com/intima/posts/144916102244400, last accessed 13/10/2018.14
Maemura, et al. 'If These Crawls Could Talk'.
14 (Meamura et al. 2018)
15 See, http://www.techcomet.com/2011/05/facebook-profiles-alternative-to.html, last accessed 13/10/2018.
16 Dragan Espenschied 'Rhizome Releases First Public Version of Webrecorder. A new Perspective in Web
Archiving', in Rhizome (2016, 9 August) https://rhizome.org/editorial/2016/aug/09/rhizome-releases-first-
public-version-of-webrecorder/, last accessed 13/10/2018.
17 However, as stated on the site some elements were adjusted: "The online exhibition version on Rhizome's
Webenact server was modified to contain contemporaneous emoji, mute links that would allow to leave the
artwork's boundary, and to not display items that have been posted before the performance started,"
https://webrecorder.io/despens/amalia-ulman-excellences--perfections, last accessed 13/10/2018.
18 Rachel Smal, 'Amalia Ulman', in: Interview (2015, 14 October) https://www.interviewmagazine.com/art/
amalia-ulman, last accessed 13/10/2018.
145