Artist Networks – An Influence Map
A Digital Arts and Humanities Approach to Art History
Table of Contents
1. Research Questions
2. Historical Perspective and Significance
3. Literature Review and Theoretical Background
4.1 Terminology and Data Collection
5. Research Journal
8. Projected Timeline
When I started the master, I already had the idea of creating visualizations of relationships between artworks and artists in some way, but I was discouraged because I have no previous experience in (a) data visualization and (b) data processing and analysis. However, I could not stop thinking about art history and how it could be included in the digital humanities to make it more accessible, especially for learners. After doing some research on possible tools for data visualization, I realized that there is a growing supply of tools for people without the necessary technical knowledge.
Thus, with my interest in learning and art history in mind, I thought about the advantages that creations and visualizations of digital networks of artworks and artists can have for learners of art history. From my personal experience, I know that studying art history is impossible without thinking of contexts, influences and relationships. However, it is often hard to grasp these complex relationships, which makes understanding art history a very difficult subject. While we have a tendency to look at artworks as individual instances, they cannot be considered without their historical, cultural, and societal context if we want to develop a deep understanding. As Paul DiMaggio writes about Becker’s work, he “argues that collaborative networks (“artworlds”) produce art and that these, rather than individual “artists”, are the proper objects of socio-scientific analysis” (2011, p.286). A lot of art history education relies on memorization of large amounts of data related to artworks in order to create a foundation on which to build one’s understanding. I argue that visual representations of relationships between artists that look at individual stories can provide great support for students of art history and facilitate learning and understanding for everyone.
I thus decided to write my dissertation on those “artworlds” mentioned by Becker and their visualization for the purpose of supporting art history education and learning. I will create a network visualization of relationship between artists and institutions for a certain period in time and analyse this digital artefact in order to draw conclusions about its usefulness, comprehensiveness, and possible applications. Questions I want to answer are: How can art history be visualized as a complex system of networks and connections in order to enable deep understanding? Which benefits can social network analysis have in the context of art history education? How can data be visualized to reflect personal stories and individual histories in the context of art history?
Until the 19th century, the art world was strictly regulated by authorities, traditions, and confined spaces, which controlled who, when and where art was to be created, published, and distributed. Since the end of the 19th century, with the decline of the academies and salons in France, and the liberation of the art market, artists had to form groups and networks in order to exhibit their work and find supporters. In their study of Impressionist art, White and White identify the Impressionists as the instigators of this change, to open up the art market for true creativity and innovation as opposed to the academies’ holding on to mass-production of traditional art (Finkelstein 1966). The academy system was too rigid for free creative expression and thus was replaced by a more dynamic and liberal system.
Of relevance for this work is the fact that with the end of the academy, there was a need for building social networks and groups, in order to support each other and promulgate art. Since the late 19th century, artists formed dynamic networks that were adaptive and flexible to react to new influences and styles. Until then, artistic epochs and periods, and connections between artists were relatively easy to identify, whereas the networks and relationships have become much more complicated since.
Especially for students of art history, it can be very hard to get a grasp of art history of modern times, with its multitude of movements, groups, collectives, and styles. This dissertation postulates that the visualization of the networks of artists since the late 19th century can be of great help for anyone trying to understand the development of art history and assist with developing a deep understanding of the organic development of artistic expression, by presenting a visual representation of artistic networks and groups and their cross-fertilization and exchange of ideas and inspiration.
Furthermore, the notion of “history” itself is in question in the postmodernist era, according to Vaughan, who writes that the “concept of history as a progression of styles orchestrated by Great Masters – has given way to a questioning of aesthetic canons and to the very notion of artistic development” (2005, p.4). University programmes of “History of Art” are now renamed “Visual Culture”, which suggests a shift from the present-centred backwards-looking traditional approach, towards “a temporally unspecific and aesthetically non-discriminatory exploration of the pictorial” (Vaughan 2005, p.4). Such a change in perspective requires old, rigid structures to be broken apart in order to make room for a more dynamic and interactive understanding of culture, including art.
Additionally, in an age of Big Data and reduction of individuals to numbers and figures, it is important to keep the individual person behind the data and their position within the social fabric of society in mind. This project aims to achieve a visualization of data about artists, that takes into account personal histories and individual stories. As Giorgia Lupi writes, we “can write rich and dense stories with data. We can educate the reader’s eye to become familiar with visual languages that convey the true depth of complex stories” (2017). The complex stories that underlie the creation of artistic works are at the focus of this dissertation. In order to understand works of art, one has to learn about their contexts and stories and the personal history of the artists who produced them. Within the discipline of art history, a lot of work is done to decipher individual artworks and “read” their iconographic and symbolic messages, but while artist’s biographies are considered, there is not much focus on social networks and personal stories of individual artists and their reflection in their work.
In the field of digital art history, Johanna Drucker’s article is a very influential piece, which reflects the ongoing struggle of digital humanities to convince art historians to renew rigid traditional methodologies. Drucker urges that there is a need to demonstrate that “digital methods change the way we understand the objects of our inquiry” (2013, p.6). She predicts that new aspects of artworks can be discovered by introducing computational, statistical and informational analysis, which “will provide new bases on which the judgment of the trained historian can build” (p.6). As an example of work in digital art history, Drucker mentions the work of Criminisi, Kemp and Zisserman (2005) on the analysis of construction of perspectival space in paintings and defines it as the basis of visual data mining. Furthermore, Drucker clearly separates digitized art history from digital art history and states that the latter is developing slowly due to the challenge that the primarily visual – as opposed to textual – object of art history pose to technology. She refers to the work of Lev Manovich in the field of cultural analytics, designed to create tools and methods for analysing large amounts of images, using parameters to sort visual information (p.8).
Manovich postulates that people “in art history need to learn the core concepts that underlie the use of data science in contemporary societies” (2015, p.15). He states that in order to create representations of data, three crucial decisions have to be made: What are the boundaries of the phenomenon, what are the objects we will represent, and what characteristics of each object will be included? (pp.15–16).
Visualizing these networks of social connections has to be approached thoughtfully as Porras points out that “[n]etwork visualizations have the potential to translate messy archival work into clouds of connection, maps of relations that can reveal hidden agents or nodes of production” (2017, p.42). I will draw on Porras specifically when identifying the challenges of this project later on.
As early as 1974, Becker’s “general proposition that knowledge and cultural product are social in character or have a social base” (1974, p.767) formed the basis for social network analysis in the field of art history. He explains that conventions are what makes the production and reception of artworks possible, and conventions are created through social interaction and negotiation. In order to understand a specific art work, Becker suggests to “think of social organization as a network of people who cooperate to produce that work” (1974, p.774).
In the field of Social Network analysis, based on Becker’s work, DiMaggio (2011) writes about art as cultural products, that are produced in certain settings and mentions four approaches that have recently been dominating network analysis of creative fields. These are “Bourdieu’s theory of competition; theories of efficient boundaries; research on small worlds; and analyses of structural mechanisms that induce creative or financial success” (p.288). DiMaggio refers to Collins’ (1989, 1998, 2000) claims about networks that drive intellectual movements, that there is a tendency of dense connections between eminent thinkers and well-known peers, which can lead to “reputational contagion” (p.289) and movement leaders creating new ties to extend their influence outside of their schools. These dense connections are the focus point of the research project and the data visualization.
Methodologically, there are four aspects of this dissertation. The theoretical aspect is the literature review and establishing a theoretical background and framework on which to build the research, as well as drawing conclusions from the visualization to tie everything together. The other aspects are practical and revolve around the digital artefact – the visualization. Firstly, data has to be collected, cleaned and stored. Secondly, the data has to be fed into the visualization software and the visualization needs to be adjusted and perfected. Additionally, every step taken towards the final visualization had to be recorded and documented in order to be published with the final artefact for transparency reasons. Finally, the visualization has to be analysed and conclusions will be drawn from what can be detected. The final discussion will look at the connections that are visualized and what can be said about them, the possible implications for art history education and learning, and possible future work in this field.
In order to be able to describe my work in an understandable manner, and keep track of my progress, it is important to establish a terminology to be used throughout the research. Lev Manovich speaks of objects and features, which make up two separate elements of data representations, giving them a modular structure (Manovich 2015, p.18). He suggests that “we adapt the term “features” to refer to both information that can be extracted through computer analysis and the already available metadata” (2015, p.16).
Hence, I will call “objects” the objects of my analysis, i.e. the artists (later on the nodes in the visualization), and “features” the metadata I extract from my data sources to describe the objects. In the data sheet, in line with the terminology used for Polinode, there are two worksheets, one called Nodes, the other Edges. The Nodes worksheet specifies the names of the nodes (e.g. Artist name) and their attributes (e.g. painter, movement). On the Edges worksheet, the source and target of the relationship between nodes are specified and the *Label attribute defines the type of relation.
As the primary source of data, I chose Wikidata (https://www.wikidata.org), which provides openly accessible metadata on the artists of interest for my dissertation. Because Wikidata does not give all desired information on all artists, I use Wikipedia (https://en.wikipedia.org) to fill in the blanks. The data for the visualization is collected as a xlsx or csv document, which can be imported into the visualization tool.
The starting point for the data gathered and visualized is the Impressionist movement, which started in the 1870s. From there on, the visualization aims at building a network as complete as possible to include the following 60 years of Western European Art history.
In the beginning of the project, I was thinking of Onodo as network visualization program, due to its user-friendliness and easy import and export of data as well as Open Source data sharing agenda, which is in line with the spirit of the Digital Humanities.
During the course of further research and getting an overview of available programs, I came across Polinode as a more suitable tool, as it offers advanced editing options. Polinode emerged in September 2015 and “is a tool for network analysis that aims to support both commercial and non-commercial use cases” (Pitts 2016, p.1424). The producers wanted to make network analysis accessible for people without the necessary background while offering advanced functionality (Pitts 2016, p.1422).
Polinode is a web-based toole that does not require desktop installations. It offers a multitude of customization options and is compatible with other tools like gephi. Furthermore, it allows markdown in order to facilitate integration into web content and it offers easy collaboration and sharing. However, unlike Onodo, Polinode is a commercial tool and no open source freeware, which I consider the downside of it.
I have created a pilot visualization of some impressionist artists with Onodo already, to give an idea of what I want it to look like, and I did the same with Polinode, using the same data sheet to experiment with the tools and find out which one is best suited for my project.
In order to keep track of my work, every step is documented in a research journal, which will be openly published with the final visualization in the end, to ensure transparency of my work and enable discussion about the process and methodology. The journal includes sources accessed, changes made, ideas and inspiration, and thoughts and critical reflection on what has already been achieved. The journal is written in notepad to facilitate export of the text to different programs.
There are various purposes of the research journal. Firstly, it helps keeping track of the progress, records the sources accessed for later reference, and gives an idea of the current state of the work. Secondly, it fosters reflection on the work itself and on the learning done so far through an ongoing process of interpretation. As Nadin and Cassell write, reflexivity “enables both in-depth thinking about the methods we use and the epistemological commitments that underlie them” (2006, p.209). Thus, the reflexive practice of keeping a research journal can itself be considered a part of the research. Thirdly, the journal provides transparency of the steps taken and the measures and tools used in order to arrive at the final product. In line with the spirit of the Digital Arts and Humanities, this research will be open and free to access for everyone in order to use, replicate and re-use it. The journal provides insight information into the process behind the final visualization and the thoughts the researcher had on the way. This information can be of great benefit for other researchers in the same or similar fields, who might come to conclusions that others oversaw.
Furthermore, given that the databases used for this research are open access and thus editable by everyone at any time, it is important to record when which data was collected as the content can vary at any time. Due to Wikidata and Wikipedia’s public histories of content and software, this can help tracing back the data that is used in case of ambiguities and errors.
One of the biggest challenges in this research is to be mindful of and identify underlying biases and political, social and cultural power differentials, as well as the impact of absent data. As Stephanie Porras writes, without methodological reflection, “network visualizations may end up reinscribing imbalances of biopolitical, cultural and social power due to the availability and assumptions of their constitutive datasets” (2017, p.42). Porras draws on Galloway’s Zuhandenheit problem “where digital tools are used unconsciously and without critical reflection” (Porras 2017, p.44; Berry and Galloway 2016) and urges that it is necessary to identify ways in which power differentials translate into the production of data (p.44). Galloway warns that “as the digital humanities expands, the ideological infrastructure will become more emboldened” (Berry and Galloway 2016, p.162), which is an important aspect to keep in mind when working in the field of digital humanities.
Concretely, the challenges of my data visualization relate to two aspects: the source of data and the display of data. Problems with the source of data include the completeness of the different sources of data and the final dataset, the possible biases and preconceptions of the collectors of the data which will be used, and, connected to that, the danger of omission by undetected absences and missing data. Problems with the display of data concern the cleaning and simplification of the final dataset, the simplification of the visualization due to issues of technological limits, legibility, and limits of time and scope.
Art history as a Western tradition has a very euro-centric viewpoint, which determines the amount and type of information that will be accessible for this project. While not all biases and preconceptions can be ruled out, the challenge is to record any instances of ambiguity, missing information, insecurity about correctness of data, and acts of simplification and cleaning of data. For this purpose, I document all my work and progress in above mentioned research journal.
In terms of the data collected and the scope of the project, there are some important limitations that have to be considered. Without a team of researchers to evaluate and rectify the collected data, errors and misrepresentations cannot be avoided. For this reason, the data collected will be limited to a few openly accessible databases, connected to the shared knowledge platform Wikipedia. This ensures the transparency in terms of the origins of the data and its generation. Regarding scope of the art history covered, this research has to be limited to a period of maximum 60 years, between the emergence of Impressionism in France in the 1870s and the end of the Weimar Republic in Germany with the rise of Hitler in 1933. Due to the geographical location of Impressionism and Expressionism in France and Germany, the focus is on this geographical area and Europe in general, however, special attention is paid to connections outside this space as well as possible relationships that would be ignored by giving too much importance to geographical boundaries.
The focus of the data collection and visualization is on individual artists rather than their works in order to limit the dataset to a manageable amount and to create a network of people and their relationships and influences, which in turn influenced their artistic expression.
Becker, H.S. 1974. Art as Collective Action. American Sociological Review, 39(6), pp.767–776.
Berry, D.M. and Galloway, A.R. 2016. A Network is a Network is a Network: Reflections on the Computational and the Societies of Control. Theory, Culture & Society, 33(4), pp.151–172.
Collins, R. 1998. The sociology of philosophies: a global theory of intellectual change. Cambridge, Mass: Belknap Press of Harvard University Press.
Collins, R. 2000. The Sociology of Philosophies: A Précis. Philosophy of the Social Sciences, 30(2), pp.157–201.
Collins, R. 1989. Toward a Theory of Intellectual Change: The Social Causes of Philosophies. Science, Technology, & Human Values, 14(2), pp.107–140.
Criminisi, A., Kemp, M. and Zisserman, A. 2005. Bringing Pictorical Spaces to Life: Computer Techniques for the Analysis of Paintings. IN: Digital art history: A Subject in Transition. Computers and the history of art (Yearbook). Bristol: Intellect.
DiMaggio, P. 2011. Chapter 20: Cultural Networks. IN: J. Scott and P. J. Carrington (eds.) The SAGE Handbook of Social Network Analysis. London; Thousand Oaks, Calif: SAGE, pp. 286–300.
Drucker, J. 2013. Is There a “Digital” Art History? Visual Resources, 29(1–2), pp.5–13.
Finkelstein, S. 1966. Reviewed Work: Canvases and Careers: Institutional Changes in the French Painting World by Harrison C. White, Cynthia A. White. Science & Society, 30(2), pp.238–241.
Lupi, G. 2017. How We Can Find Ourselves in Data.
Manovich, L. 2015. Data Science and Digital Art History. International Journal for Digital Art History, 1(1), pp.13–35.
Nadin, S. and Cassell, C. 2006. The Use of a Research Diary as a Tool for Reflexive Practice: Some Reflections from Management Research. Qualitative Research in Accounting & Management, 3(3), pp.208–217.
Pitts, A. 2016. Polinode: A Web Application for the Collection and Analysis of Network Data. IN: 2016 IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM). ASONAM. San Francisco, CA: IEEE, pp. 1422–1425.
Porras, S. 2017. Keeping Our Eyes Open: Visualizing Networks and Art History. Artl@s Bulletin, 6(3), pp.42–49.
Vaughan, W. 2005. History of Art in the Digital Age: Problems and Possibilities. IN: A. Bentkowska-Kafel, T. Cashen, and H. Gardiner (eds.) Digital Art History: A Subject in Transition. Computers and the History of Art. Bristol, UK ; Portland, OR: Intellect, pp. 3–13.