Expert/Crowdsourcing for the Linked Modernisms Project

Scholarly and Research
Communication

Stephen Ross, Alex Christie, & Jentery Sayers

volume

5/

issue

4 / 2014

University of Victoria, Modernist Versions Project

Introduction
ough “Building Partnerships to Transform Scholarly Publishing” has never oﬃcially
been part of the mandate of either the Modernist Versions Project (MVP) or Linked
Modernisms (LiMo), upon reﬂection, it has turned out to be among their key
functions. Advancing new ways of doing scholarship within the ﬁeld of modernist
studies has driven us to explore novel approaches to publishing, and demonstrated the
utter necessity of partnerships to make those approaches eﬀective. In what follows we
will provide a quick introduction to the Modernist Versions Project1 and some of the
tools we use, consider what sort of inquiries appear at scale, and give an overview of
the Linked Modernisms Project. We will attend in particular to how this suite of
projects has led us to rethink traditional scholarly publishing practices, and how
partnerships have become indispensable to mobilizing our research.

CCSP Press
Scholarly and Research Communication
Volume 5, Issue 4, Article ID 0401186, 11 pages
Journal URL: www.src-online.ca
Received July 31, 2014, Accepted August 25, 2014, Published December 16, 2014

Stephen Ross is Associate
Professor of English and
Cultural, Social, and Political
ought at the University of
Victoria. He is author of
Conrad and Empire (2004);
editor of Modernism and
eory (2009); co-editor of
e Modernist World (2015),
Dorothy Richardson’s Pointed
Roofs, and Dorothy
Richardson’s e Tunnel; and
General Editor of the
Routledge Encyclopedia of
Modernism. He is Principal
Investigator of both the
Modernist Versions Project
and Linked Modernisms.
Email: saross@uvic.ca
Alex Christie is a doctoral
research assistant with
Implementing New
Knowledge Networks (INKE)
and the Modernist Versions
Project (MVP). He works
across the Electronic Textual
Cultures Lab (ECTL) and the
Maker Lab in the Humanities
at the University of Victoria,
using new and built media to
model the procedural nature
of modernist texts. Email:
achris@uvic.ca

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.
© 2014 Stephen Ross, Alex Christie, & Jentery Sayers. is Open Access article is distributed under
the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons
.org/licenses/by-nc-nd/2.5/ca), which permits unrestricted non-commercial use, distribution, and
reproduction in any medium, provided the original work is properly cited.
1

Scholarly and Research
Communication
volume

5/

issue

4 / 2014

Jentery Sayers is Assistant
Professor of English and
Director of the Maker Lab in
the Humanities at the
University of Victoria. His
work has appeared in
American Literature; e-Media
Studies; Digital Studies / Le
champ numerique; e
International Journal of
Learning and Media; Kairos;
Rhetoric, Technology, and
Pedagogy; e Information
Society; and Computational
Culture, among others. Email:
jentery@uvic.ca

The modernist versions project
e MVP sought to help transform scholarly publishing through partnerships from the
start. Funded by a Partnership Development Grant through the Social Sciences and
Humanities Research Council (SSHRC), it has played a role in the increasing practice of
putting scholarly edited and curated materials online as open access digital editions.
Our most prominent publication of this nature is the ﬁrst edition of James Joyce’s
Ulysses, which we published over the course of a year from June of 2012–June of 2013
(Joyce, 2012). We supplemented the presentation of PDF page images and OCRcorrected text with a series of Twitter chats led by international Joyce scholars and a
series of video podcasts of lectures by prominent Joyceans2. From the outset, the
mindset was that of leveraging the partnerships upon which the MVP was founded – in
this case the Modernist Journals Project, Fairleigh Dickinson University, the Electronic
Textual Cultures Laboratory (ETCL), and Implementing New Knowledge Environments
(INKE) – to help rethink what it would mean to present the ﬁrst edition of Ulysses in a
scholarly setting that was equally compatible with a public humanities ethos.
e production of digital editions was in the ﬁrst instance a matter of necessity for the
MVP as it sought to pursue its more focused goal of digitally versioning texts. In the
spirit of genetic criticism’s pursuit of the avant-texte3, the MVP again ventured into
unfamiliar scholarly publishing terrain. As Stephen Ross (2012) wrote early on: “e aim
is to read the gap between one version and the next, following Derrida’s imperative to
attend to diﬀerence itself as the source of meaning rather than presuming that it is a
merely negative product of the non-coincidence of two entities: text A and text B” (n.p.).
We sought to publish the diﬀerence between two existing witnesses, and to say what that
diﬀerence meant. In one sense, this is just literary criticism as it has long been practiced.
In another sense, it nuances the notion of the avant-texte signiﬁcantly by showing how
diﬀerential texts produce “ghost texts” that are perhaps even more signiﬁcant than those
tied to material artifacts. Pursuit of this admittedly evasive objective required that the
MVP build its partnerships with Editing Modernism in Canada, Islandora and
DiscoveryGarden, and Juxta (through NINES). Doing so allowed us to draw upon
enormous expertise and deep experience with some of the same issues we were just
embarking on. It also gave us access to state-of-the-art digital collation and versioning
applications with which to experiment as we developed various workﬂows for
generating new critical insights via digital collation of texts. We experimented with
conceptual versioning and various levels of Text Encoding Initiative (TEI) encoding
both before and aer collating texts (e.g., Tanigawa, 2013a, 2013b, 2014).
Over time, though, our focus shied away from textual variation per se and onto other
ways we might conceive of “modernist versions.” Most recently, this shi has taken two
decisive forms. First, Jana Millar Usiskin, working through our partnership with
Compute Canada, has built a repository of modernist novels for topic modelling and
other statistical modes of inquiry, as well as producing network visualizations of some
key modernist texts (see Millar Usiskin & Moa, 2013, 2014a, 2014b). Adapting Matt
Jockers’ (2013a, 2013b) work with Victorian novels to the diﬀerent demands of
modernist texts, Millar Usiskin and Moa have reﬁned Jockers’ recommended workﬂow
to conform more accurately to modernist novels’ preoccupation with verbs, sentiment,
and time. Second, Katie Tanigawa, Alexander Christie, and Adèle Barclay, working
closely with Jentery Sayers, have developed what we are calling Z-Axis Scholarship.

2

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

is approach involves geo-referencing (with TEI) modernist novels and then using
the resulting data to warp historical maps of cities such as Paris and Dublin (see
Barclay, 2014; Christie, Ross, Sayers, Tanigawa, & the INKE-MVP Research Team,
2014). e results, though still preliminary, have been enticing, suggesting some new
lines of inquiry regarding class, religion, and sexuality in Ulysses, Nightwood by Djuna
Barnes, and Quartet by Jean Rhys.

Scholarly and Research
Communication
volume

Finally, working with these tools and partners helped us discover what sorts of
intervention we needed that we would have to build for ourselves, provoking an
editorial and critical enterprise like the MVP to take on the unwonted task of
designing and building soware. Enter Susan Schreibman and Tanya Clement’s
Versioning Machine (Figure 1). Having identiﬁed what worked for us – and what did
not – through our experiments with a number of other tools (outlined in Huculak &
the MVP Research Team, 2011), and with the invaluable creative and technical
contributions of MVP team member Daniel Carter, the Tools Team of the MVP has
now produced Versioning Machine 5.0 (Carter, 2014), which is due for release in beta
in the fall of 2014. e new Versioning Machine presents an intuitive user interface and
plentiful features for versioning work in both text and oral formats, as well as across
formats.
Figure 1: Screen capture of the Versioning Machine 5.0

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

3

5/

issue

4 / 2014

Scholarly and Research
Communication
volume

5/

issue

4 / 2014

Looking ahead, the MVP will continue all three lines of work as it develops existing
partnerships and evolves new ones to continue exploring innovative approaches to
scholarly publishing.

Inquiries appearing at scale
ough the MVP began with a narrow interest in how multiple witnesses of modernist
texts varied among themselves, it has quickly embraced Alan Liu’s (2012) claim that
“Scale is a new horizon of intellectual inquiry” and his follow-up question: “What kinds
of humanistic phenomena appear only at scale?” (p. 21). Until very recently, the MVP
lacked a suﬃciently large body of clean or reliable texts to work with. Largely for reasons
of copyright, scholars have been slow to digitize modernist texts. As we began building
our own repositories, then, we also began exploring the larger possibilities of Linked
Open Data. is precipitated a shi in our objects of inquiry, away from individual texts
or avant-textes and toward the much larger world of modernist cultural production. To
ensure suﬃcient complexity, we also wanted to consider how that world itself is
produced as a ﬁeld of study. We began to consider how we could productively study
both individual texts and the much larger ﬁeld of modernist cultural production as
networks that could usefully be visualized and explored with digital methods. By
eavesdropping on the INKE team’s developing paper on “e Emergence of Linked Data
in the Humanities” (Simpson et al, n.d.) and investigating projects such as linkedjazz.org
and europeana.eu, we quickly realized the potential aﬀorded by Resource Description
Framework (RDF) and other means of faceted searching of linked databases. When we
came across the work of Colin Allen, Mathias Niepert, and Cameron Buckner on the
Indiana Philosophy Ontology (Allen, Niepert, & Buckner, 2013), we knew we had found
the ideal precedent and model for our own work.
e Indiana Philosophy Ontology (InPhO) uses the open access Stanford Encyclopedia
of Philosophy (SEP) as its database, and has devised various approaches to making the
material in it searchable and discoverable in sophisticated ways. is has involved
constructing an information ontology for the terms and entities in the SEP, and
devising methods of dealing with dynamic ontologies as a way of responding to an
ever-growing and shiing body of knowledge. at is, as the SEP is updated and
expanded, the ontology that allows for its navigation must also expand and change,
retroactively establishing connections with terms that derive from diﬀerent moments
in the SEP’s production. Additionally, the InPhO researchers have been able to use
natural language processing (NLP) and a combination of machine learning with
human domain expertise checking to automate the process of updating the dynamic
ontology as much as possible.
ese approaches, combining intensive processing capacity with light-touch human
intervention, have been extremely productive. ey have allowed experts to correct or
reﬁne the machine ﬁndings, and they have led experts to new insights based on how
machine inferences connect terms. As Buckner, Niepert, and Allen (2010) put it, such
approaches are crucial to the ongoing viability of encyclopedic enterprises in the
twenty-ﬁrst century: “Scholars and students don’t just need the reference works – they
also need the means to search and navigate them eﬀectively. To preserve the utility of
encyclopedias as they grow, we must also improve our ability to represent their
contents in meaningful ways accessible to novice and expert alike” (n.p.).
4

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

Perhaps even more to the point of the concerns presented here, InPhO have pioneered
a hybrid model of scholarly publishing that we believe to be exemplary for digital
humanities projects. ey have taken a commercially published resource (SEP) and
made it open access through academic investment. ey then partnered with Indiana
University to support the ongoing development and reﬁnement of InPhO’s ontology
and faceted browsing capabilities. Central to this project have been RDF, NLP, and
inferential algorithms of precisely the sort required by Linked Modernisms as it
develops a similar project in relation to the Routledge Encyclopedia of Modernism
(REM).

Scholarly and Research
Communication
volume

Linked modernisms
Linked Modernisms emerged from thinking about how we could best leverage RDF
and faceted searching for big data inquiries into modernist cultural production. Linked
Modernisms aims to allow researchers to discover (perhaps serendipitously)
relationships among versions of modernism; visualize and plumb those versions; reﬁne
queries across subjects, objects, and predicates; and develop nuanced understandings
across disciplinary, artistic, temporal, linguistic, or geographical articulations of
modernism. In eﬀect, LiMo seeks to be able to ask questions of and characterize
relationships across all conceivable entities associated with modernist cultural
production wherever, whenever, and however they occur. It is a tall order.
To ﬁll it, LiMo envisions a four-tier information matrix that also describes a new
approach to scholarly publishing through a range of partnerships. First, information
included in the REM itself; second, metadata about the terms in the REM provided to
us by domain experts; third, metadata gathered through NLP and inferential
algorithms; and fourth, metadata about those who provided both the initial data (the
entries) and the ﬁrst round of metadata. ough no one would claim such a matrix to
be complete, we will have at our disposal both the most expansive and inclusive body
of knowledge about modernism and the production of the ﬁeld of modernist studies as
well as a sophisticated means of analyzing that data. e resulting information matrix
emphasizes discoverability, allowing for serendipitous encounters with the datasets,
queries driven by user curiosity, and user-generated itineraries through the datasets. It
presents an innovative approach to scholarly publishing inasmuch as it fuses a
proprietary dataset published by a commercial press with an open access metadataset
that is partially derived from the processing of the proprietary materials. Both datasets
are analyzed by grant-funded research projects in Canada and the USA, with the
results appearing in both traditional peer-reviewed and non-traditional open access
fora. e whole can be achieved only through innovative partnerships among
academic researchers, institutional units, and commercial interests.
Tier I of LiMo’s information matrix is the REM itself, produced under contract for
Routledge by General Editor Stephen Ross with the understanding that while copyright
over the contents remains with the press, they may be used by Ross for nonconsumptive or non-expressive purposes (a factor that will become especially
signiﬁcant in using NLP to infer relationships among terms in the data and metadata).
Moreover, though Routledge owns the copyright to the content, they do not own any of
the metadata we harvest by other means. We have agreed to share our metadata with

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

5

5/

issue

4 / 2014

Scholarly and Research
Communication
volume

5/

issue

Routledge in the spirit of open access, but have insisted that the metadata must remain
open access.

4 / 2014

With eight Subject Editors, 60 Editorial Board Members, and approximately 1,200
contributors, the REM aims to include everything that may be called modern(ist)
across the arts and around the world, with little concern for strict periodization or
narrow aesthetic criteria. It features nearly 3,000 terms and runs to approximately 1.5
million words, all of them written by domain experts. It constitutes a new venture for
Routledge itself, which has taken on the task of producing a large online resource of
curated scholarly material rather than contracting it out (as with the Routledge
Encyclopedia of Philosophy). When the REM launches in 2016, it will be the most
expansive source of information on modernism in the world. It is the bedrock of
Linked Modernisms.
Tier II of Linked Modernisms gathers metadata directly from domain experts through
a survey. Each term in the REM has abundant implicit metadata; it carries with it an
information ecology that may or may not be embedded in the entry itself, and which
aﬀords a rich source of data for Linked Modernisms’ Linked Open Data (LOD)
program. Most of this data, however, is not recorded in the entries themselves, and was
not at the time routinely collected by Routledge4. e ﬁrst order of business for Linked
Modernisms was to determine what metadata we would ideally like to gather (all of it),
versus how much we could reasonably expect our contributors to provide. is entailed
taking the ﬁrst steps toward building an ontology that would encompass all aspects of a
wide range of entities, so we established a core framework with fundamental classes
and sub-classes of entity. In this respect, LiMo faced a challenge larger than that
presented to either the designers of linkedjazz.org or InPhO: where they focus only on
one subject or discipline, LiMo had to be able to accommodate a range of disciplines,
with a range of entities – including people, movements, techniques, works and artifacts,
locations, languages, and events. Our framework seeks to capture any term related to
modernism in dance, ﬁlm, literature, theatre, music, architecture, design, visual arts,
and intellectual currents – anywhere around the world. As we transform this core
framework into an oﬃcial ontology, we will use existing ontologies as much as possible
– only devising our own features where necessary – to preserve maximum
interoperability with other projects and to ensure maximum usefulness in the LOD
environment. We plan to publish the ontology in early 2015 as both a useful resource
and a scholarly contribution to taxonomizing the ﬁeld of global modernism.
With the framework in place, we built a survey on SurveyMonkey5 designed to harvest
the relevant metadata about the entries in the REM from the contributors themselves.
at is, we are crowdsourcing information from domain experts, achieving a level of
accuracy and reliability not normally associated with crowdsourcing, and a breadth of
information not normally associated with reliance upon domain experts. e survey is
currently in circulation with over 1,200 contributors to the REM.
e results of the survey, distributed using SurveyMonkey, are human-readable and
allow us to gain an impressionistic overview of the metadata breakdown. ey are
exportable in spreadsheets, CSV, XML, HTML, summary form, or question by question.
We have had the most success in exporting the results in CSV and using a free online
6

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

converter to produce JSON objects from the results. e JSON objects must be cleaned
up either manually or by running a simple Python script to automate, before they can
be loaded into a triple-store such as Apache Jena or Neo4J, which can then be queried
using SPARQL. When thus exported into RDF, the survey results – the metadata for the
REM – assume a machine-readable format that enables more complex analysis,
including options for faceted searching, network visualizations, and querying degrees
of separation between entities in the domain.

Scholarly and Research
Communication
volume

e key enabling fact in terms of innovations in scholarly publishing and new models
of partnership is that though the contents of the REM are under copyright by Routledge,
the metadata and survey results are not. As such, both the raw metadata and the
analyzed results are open access and public. We note with some wry (dis-)satisfaction
that Routledge appears to have learned from this experience, and perhaps others like it:
at least some contributor contracts issued by Routledge now include a clause requiring
authors to complete a metadata survey in addition to providing the content for which
they have been contracted.
Tier III takes Linked Modernisms into the realm of Natural Language Processing
(NLP) and algorithmic harvesting of metadata. With our data output in machinereadable RDF, Linked Modernisms will trigger a partnership with InPhO to draw upon
their expertise in automated and semiautomated data analysis. Colin Allen, Mathias
Niepert, and Cameron Buckner are leaders in using NLP, algorithmic identiﬁcation of
collocated terms and topic modelling, and human-machine interaction to ensure highquality results. Our collaboration with them will include analyzing both the results of
the metadata survey and the content of the REM. rough our partnership we will
follow the protocols devised by InPhO in analyzing the content of the Stanford
Encyclopedia of Philosophy. e academic partnership between Linked Modernisms
and InPhO, drawing upon material – some of which is copyright protected, some of
which is not – produced for a for-proﬁt publisher will help chart new territory in
scholarly publishing.
Equally, Linked Modernisms will draw upon its partnership with Susan Brown’s
Canadian Writing Research Collaboratory (CWRC) and the Orlando Project6 (Brown,
Clements, & Grundy, 2010) to pool resources and eﬀorts. For some time, the CWRC
and Orlando have been pursuing many of the same methodological questions that
Linked Modernisms is now engaging, with signiﬁcantly overlapping material.
Continuing the theme of developing partnerships to transform scholarly publishing,
this collaboration will be informed by InPhO’s contributions and will aim to establish
interactive forms of scholarly publishing that let researchers beneﬁt from copyrighted
material without necessarily having access to that material. at is, where copyright
restricts the material produced by the Orlando Project for Cambridge University Press
and by the REM for Routledge, metadata surveys and NLP approaches to the content –
non-consumptive or non-expressive uses – allow researchers to analyze materials
without necessarily being able to read them directly. In one respect, this may be one of
the most signiﬁcant beneﬁts to distant reading, as proprietary repositories can still be
used to yield scholarly insights even when they remain oﬀ-limits to researchers, or at
least diﬃcult for them to access.

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

7

5/

issue

4 / 2014

Scholarly and Research
Communication
volume

5/

issue

4 / 2014

One area in which Linked Modernisms will collaborate most fruitfully with CWRC,
Orlando, and INKE is the relatively new (to humanities) terrain of sophisticated
visualizations of research results. e manifold partnership of these projects with
InPhO’s advanced algorithmic tools will help develop John Saklofske’s (2013)
NewRadial visualization application as a form of publishing. at is, when researchers
execute queries on data that yield signiﬁcant results, they will be able to publish them
in scholarly form as visualizations that demonstrate graphically the arguments they are
pursuing. is may in fact be the most exciting aspect of the kind of work being done
by the Orlando Project, INKE, and Linked Modernisms, as it aﬀords scholars not just
new ways to understand and present their research results, but innovative ways to
query their ﬁelds.
Tier IV of the Linked Modernisms information matrix takes a self-aware approach to
the very construction of the ﬁeld of modernist studies. Supplementing the content
(REM), ﬁrst level of metadata (survey), and second level of metadata (NLP and
algorithmically produced) of Linked Modernisms is a third level of metadata. It
concerns how, by whom, and with what invisible biases the ﬁeld of global modernist
studies (as represented by the REM and its associated metadata) is constructed.
Proceeding from the knowledge that no ﬁeld of knowledge is neutrally constructed, we
have constantly wondered how we can make legible the invisible or latent factors
governing the production of the ﬁeld of global modernist studies as the REM and
Linked Modernisms will capture it. at is, if the knowledge contained in the REM’s
entries and the metadata gathered about those entries comes from particular
individuals – even if they are domain experts – then how can we trace the particularity
of those individuals? How, that is, can we prevent the ﬁeld of knowledge from posing
as neutral and merely factual? How can we reconstruct the argument latent in that ﬁeld
of knowledge?
In an eﬀort to answer these questions, we have devised a second information ontology
framework that seeks to capture all relevant metadata about the individuals who
provided the metadata in the ﬁrst survey (meta-metadata?). is framework has
become the basis of another survey that asks demographic questions of the domain
experts with an eye to capturing an overview of those characteristics that most oen
remain unmarked, even though they can profoundly aﬀect the information proﬀered:
sex, gender, sexuality, ethnicity, religion, language, location, professional status, training,
etc. We have received ongoing Human Research Ethics Board approval 7 to conduct this
research, and intend to implement the survey in the coming year. As with the ﬁrst two
levels of metadata derived from the REM and its ﬁrst survey results, this survey will
also be supplemented by NLP and algorithmic processing to infer relationships that
may otherwise remain obscure. at is, even where respondents believe themselves to
be completely honest about their backgrounds and partiality, inferential algorithms
may be able to detect further occulted aspects of how they produce and present
information, which will be signiﬁcant for understanding modernist studies as itself a
ﬁeld of data produced in speciﬁc circumstances.

Hybrid publishing and partnerships
In terms of the present publication, perhaps the most interesting element of Linked
Modernisms is its reliance upon fusing multiple publication models through
8

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

partnerships with a commercial press and other academic initiatives. e REM itself is
a commercial publishing venture by Routledge on behalf of the global corporation
Taylor and Francis. Linked Modernisms will harvest metadata based upon the content
of the REM. Simultaneously, the metadata will feed into the faceted searching
apparatus being devised by Routledge as the primary element of value added to the
REM, setting it apart from other online encyclopedias. e metadata will be gathered
for open access viewing and use by Linked Modernisms, which is housed at the
University of Victoria. It will be published online through mvp.uvic.ca – an academic
domain. e analysis of the metadata will be done by Linked Modernisms in
collaboration with InPhO, housed at the University of Indiana, and CWRC and the
Orlando Project, housed at Guelph University and the University of Alberta,
respectively. e results will be disseminated in a further hybrid format, with raw and
preliminary results openly accessible online, and with more in-depth analyses
submitted for peer-reviewed publication in both online and print formats. Moreover, as
the REM expands and undergoes revision, so Linked Modernisms will continue to
gather metadata and analyze it in partnership with InPhO and Orlando, creating a
recursive publishing environment in which those contributing to the REM aer the
results of Linked Modernisms are made public may be doing so in light of – or at least
with an awareness of – the initial ﬁndings. is awareness will add a layer of selfreﬂexivity to the production of knowledge about global modernist studies and the ﬁeld
of modernist studies itself, which can in turn be analyzed as part of the ongoing
analysis of the metadata.

Scholarly and Research
Communication
volume

As Linked Modernisms matures, we remain keen to develop modes of scholarly
publishing and dissemination that are suited to the dynamic nature of the knowledge
ﬁeld under consideration – both data and metadata – and that conform to the
principles of open access and creative commons sharing. We welcome any suggestions
along these, or indeed any other, lines.

Notes
1. e MVP would like to acknowledge and thank its partners, without whose aid
very little of what is described here would have been possible: Networked
Infrastructure Nineteenth-century Electronic Scholarship (and Juxta), the
Modernist Journals Project, Islandora/DiscoveryGarden, Editing Modernism in
Canada, Implementing New Knowledge Environments, the Humanities Computing
and Media Centre at the University of Victoria (UVic), and Fairleigh Dickinson
University. We are grateful to the Social Sciences and Humanities Research Council
of Canada (SSHRC) and the Maker Lab in the Humanities at UVic for their support
of this project, and to the Electronic Textual Cultures Lab at UVic, and Ray
Siemens, in particular, for facilitating it.
2. A summary of this activity is now published online on the University of Victoria
Library’s site, UVicSpace (see Krecsy & Modernist Versions Project Team, 2014).
3. is concept assumes that all materials from notes through manuscripts,
typescripts, corrections, proofs, rejected versions, deletions, additions, all print
editions, and any corrected or collected editions – taken together – constitute the
literary object of study. e avant-texte is a textual ecosystem in which the oﬃcial,
Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

9

5/

issue

4 / 2014

Scholarly and Research
Communication
volume

5/

issue

ﬁrst, ﬁnal, or approved edition is but one element, held to be of no greater
importance than any other.

4 / 2014

4. is has apparently changed, as Routledge now circulate a metadata survey to all
contributors to their publications.
5. Our preference now is to use Fluid Surveys or another Canadian-based company,
due to concerns about data security with American companies, but it was
prohibitively complicated to migrate the survey from one platform to the other, so
we continue to use SurveyMonkey for this project.
6. Like Linked Modernisms, the Orlando Project aims for open access but is
constrained by its relationship with Cambridge University Press, which owns the
copyright to its material.
7. HREB Approval Certiﬁcate 12-372.

References
Allen, Colin, Niepert, Mathias, & Buckner, Cameron. (2013). The Indiana philosophy ontology project.
URL: https://inpho.cogs.indiana.edu/ [July 31, 2014].
Barclay, Adèle. (2014). Reading and georeferencing Rhys (Sayers, Jentery, Wilson, Karly, & the Maker
Lab in the Humanities, Eds). URL: http://web.uvic.ca/~mvp1922/reading-and-georeferencingrhys/ [July 31, 2014].
Brown, Susan. (2009). The Canadian writing research collaboratory/Le Collaboratoire scientiﬁque des
écrits du Canada. URL: http://www.cwrc.ca/projects/infrastructureprojects/pilot-projects
/women-and-words-canadian-feminist-literary-collectives-1980-2000/ [July 31, 2014].
Brown, Susan, Clements, Patricia, & Grundy, Isobel. (2010). The Orlando project. URL: http://www
.artsrn.ualberta.ca/orlando/ [July 31, 2014].
Buckner, Cameron, Mathias Niepert, & Colin Allen. (2010). From encyclopedia to ontology: Toward
dynamic representation of the discipline of philosophy. Synthèse. URL: https://inpho.cogs.indiana
.edu/papers/2010BucknerNiepertAllen.pdf [July 31 2014].
Carter, Daniel. (2014). The versioning machine for audio: Introducing VM 5.0. The Modernist
Versions Project. URL: http://web.uvic.ca/~mvp1922/the-versioning-machine-for-audiointroducing-vm-5-0/ [July 31, 2014].
Christie, Alex, Ross, Stephen, Sayers, Jentery, Tanigawa, Katie, & the INKE-MVP Research Team.
(2014). Z-Axis Scholarship: Modeling how modernists wrote the city. The Modernist Versions
Project. URL: http://web.uvic.ca/~mvp1922/z-axis/ [July 31, 2014].
Huculak, J. Matthew, & the MVP Research Team. (2011). MVP White paper on collation tools. The
Modernist Versions Project. URL: http://web.uvic.ca/~mvp1922/mvp-white-paper-on-collationtools-2011/ [July 31, 2014].
Jockers, Matt. (2013a). Macroanalysis: Digital methods and literary history. Champaign, IL: University
of Illinois Press.
Jockers, Matt. (2013b). “Secret” recipe for topic modelling themes. URL: http://www.matthewjockers.net
/2013/04/12/secret-recipe-for-topic-modeling-themes/ [July 31, 2014].
Joyce, James. Ulysses. (2012). The ﬁrst edition. The Modernist Versions Project (Huculak, Matthew J.,
Curator). URL: http://web.uvic.ca/~mvp1922/portfolio/ulysses-shakespeare-co-1922-1st-edn/
[July 31, 2014].

10

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

Krecsy, Stefan, & Modernist Versions Project Team. (2014). Year of Ulysses. URL: https://dspace
.library.uvic.ca/handle/1828/5484 [July 31, 2014].
Liu, Alan. (2014). The state of the digital hmanities: A report and a critique. Arts and Humanities in

Scholarly and Research
Communication
volume

Higher Education,11(8), 8-42.
Millar Usiskin, Jana, & Belaid Moa. (2014a). Making models of modernism. The Modernist Versions
Project. URL: http://web.uvic.ca/~mvp1922/mmm/ [July 31, 2014].
Millar Usiskin, Jana, & Belaid Moa. (2014b). Visualising communities in Mrs. Dalloway. The Modernist
Versions Project. URL: http://web.uvic.ca/~mvp1922/dallowaycommunities/ [July 31, 2014].
Millar Usiskin, Jana, & Belaid Moa. (2013). Making modernism big. The Modernist Versions Project.
URL: http://web.uvic.ca/~mvp1922/big/ [July 31, 2014].
Ross, Stephen. (2012). Theory of versioning. The Modernist Versions Project. URL: http://web.uvic.ca
/~mvp1922/theory-of-versioning/ [July 31, 2014].
Saklofske, Jon, & the INKE Modelling and Prototyping Team. (2013). NewRadial. http://inke.acadiau
.ca/newradial/ [July 31, 2014].
Simpson, John, Brown, Susan, Sayers, Jentery, Bath, Jon, Quamen, Harvey, Saklofske, Jon, Barclay,
Adèle, Christie, Alex, Elliott, Mandy, & the INKE Research Group. The emergence of linked data in
the humanities. [Unpublished paper]
Tanigawa, Katie. (2014). Visualizing a changing Nostromo. The Modernist Versions Project. URL:
http://web.uvic.ca/~mvp1922/visualizing-a-changing-nostromo/ [July 31, 2014].
Tanigawa, Katie. (2013a). Returning to the periodical context. The Modernist Versions Project. URL:
http://web.uvic.ca/~mvp1922/returning-to-the-periodical-context/ [July 31, 2014].
Tanigawa, Katie. (2013b). Warping the city: Joyce in a mudbox. Maker Lab in the Humanities. URL:
http://maker.uvic.ca/mudbox/ [July 31, 2014].

Ross, Stephen, Christie, Alex, & Sayers, Jentery. (2014). Expert/Crowdsourcing for the Linked
Modernisms Project. Scholarly and Research Communication, 5(4): 0401186, 11 pp.

11

5/

issue

4 / 2014