From Network to Patchwork Collections

Canberra.
And we’re starting the final day of the Digital Humanities Australasia 2012 conference. The day begins with a keynote by Julia Flanders, who challenges us to rethink collections. This begins by asking what we mean by a ‘collection’, in the first place: collection implies agency, a collector who creates a sense of order amongst the entities they collect.

There’s a bounded comprehensiveness which is implied by the term, too – a completeness may have been achieved through the collection. A collection is an aggregation of individual items which are not just meaningful in themselves, but importantly also in their relationships.

Digital collections may also be described as initiatives, archives, projects, or libraries, for example, and each of these terms implies further connotations. They don’t express the genre of the work being created, but point to certain preoccupations – and more recently, have started to converge in a set of assumptions which don’t match the paradigm of an ‘edition’ or ‘archive’.

Collections also tend to imply that their data and metadata will be modelled at some point: that there is an aim to gain a purchase on their internal semantics, to represent such semantic data in a meaningful, useful way – through a user interface of some form.

Collections are deliberate aggregations that function as an ecology; their interfaces have become standard and familiar, and the recent attention on the value of scale reinforces the naturalness of this aggregating approach as an affordance of the digital medium. But this is distinct from the idea of simple totality: there is deliberate scope and deliberate shaping, rather than a craving for ‘everythingness’. Primary materials are common in this context, but this need not necessarily be so.

Julia outlines this in the context of a number of collections she’s involved with; there are underlying collection design similarities even though they deal with very different materials and have very different intent.

So how do we constitute a collection? There’s a break between a then and a now: material, offline collections were determined by a geographical proximity, but for online, digital collections this is no longer true; this points to a greater flexibility of organisation, of the information architecture, and potentially also enables users to gain a greater agency in imposing their own information structures.

What emerges from this is a philosophical urgency as a driving force behind the design of digital collections: it considers the standards of intellectual adequacy which a collection must meet. The design logic behind such collections, working in the formats in which they do, is a craving for more information, more context, more richness, but of course still constrained by the specific formats chosen.

What if we think of these bounds as intellectually enabling, and to do with genre rather than with technological limitations? If we identify a philosophical urgency behind these questions, we can also ascribe a certain agency to it.

What are we modelling in the design of the collection, then? First of all, agency in the development of the collection (creation, sponsorship, curation, readership); the design of the collection as a collection (coding methods, interface design, etc.); the identities, boundaries, and contents of the entities it contains; the collection’s degree of internal homo- or heterogeneity; the explicit interconnections between items in the collection; the explicit collection-level phenomena in the collection (through predefined views by timeframe or subject, for example); the implicit collection-level phenomena inherent in the collection (commonalities and patterns which are revealed in the process of examination, through datamining); and the reader’s epistemological relationship to the collection (the in-built assumptions about the reader’s existing knowledge).

How does such modelling affect us as readers? The modelling takes us from our conscious actions as collection creators to the wider ecology of the collection, and the readers’ possible actions within it. Searching produces a contingent, just-in-time, dynamically generated collection, for example; it infers the presence of an answer from a set of markers to the answer. Browsing, on the other hand, builds on the organisation of the collection itself, its collectedness; the collection frames and usefully constrains the browsing and thereby points the reader in certain directions.

And of course there can be transitions between those approaches: searching can constrain the collection to matching records, which may then be browsed and explored. But this can also happen through data mining; mining algorithms are themselves collection models.

What is the cultural significance of collections, then? We can think of the network as a flow of information via interconnection; by contrast, the patchwork is an assemblage, a suturing together of previously unrelated, but connectable elements. The network models things which are there to be modelled; the patchwork asserts juxtapositions by force.

How does this translate to collections? What would a collection based on patchwork rather than network look like? We must think about collections as an activity: a network is underpinned by a set of communications, and and a networked connection assumes from the start, optimistically, that implicit connections are there to be discovered, marked, made explicit.

A patchwork connects preexisting informational nodes that are acknowledged to be incommensurable, that are the wrong size and shape and reshapes them, makes connections, fits them together under a new design. A patchwork is cobbled together ad hoc, hacked, in a downright scandalous way; it accepts friction and imperfection, rather than assuming orderedness, conformity, and compatibility.

The patchwork collection can also express and support analysis, acknowledging the effects of its explicit patchworking on the collection itself. It permits a distinctively important kind of intellectual transaction, without the apparent passivity of the library, but with a negotiation of meaning which acknowledges the agency of the collection itself as well as the agency of the reader or user.