Periodic Reporting for period 3 - ImagiDem (Imagi(ni)ng Democracy: European youth becoming citizens by visual participation)
Reporting period: 2021-07-01 to 2022-12-31
The public sphere of today's youth is increasingly dominated by visual content. Therefore, the current youth's understanding of political action – building arguments, mobilizing, and participating – is also likely to be and increasingly become anchored in repertoires of visual participation and political action.
The research on democratic practices and political participation has thus far concentrated heavily on words. Now, perhaps more than ever, understanding political participation on non-verbal levels has become equally crucial in order to build conceptual understanding of the requirements of democratic inclusion of all citizens and find remedies to the crises in democratic governance.
Future demands for European democracy will be voiced by European youth of the 2010s and 2020s. This generation builds its democratic imagination and views of political participation in a polarized and turbulent political climate. What kind of citizens will they be, and are they already? How will and do they shape the political culture and institutions of European democracy? What are the means of participation they rely on, and how do they engage in processes of politicization? What is – and what will be – the meaning of politics and the value of democracy for them?
In ImagiDem, we study visual participation both online and offline. We analyse images, stories and memes posted on social media, and follow young people's visual ways of participation as part of their everyday actions. The project combines visual ethnography with computational big data minining and analysis, and deploys this combination to a comparative research setting in four European countries: Finland, France, Germany, and Portugal.
The research on democratic practices and political participation has thus far concentrated heavily on words. Now, perhaps more than ever, understanding political participation on non-verbal levels has become equally crucial in order to build conceptual understanding of the requirements of democratic inclusion of all citizens and find remedies to the crises in democratic governance.
Future demands for European democracy will be voiced by European youth of the 2010s and 2020s. This generation builds its democratic imagination and views of political participation in a polarized and turbulent political climate. What kind of citizens will they be, and are they already? How will and do they shape the political culture and institutions of European democracy? What are the means of participation they rely on, and how do they engage in processes of politicization? What is – and what will be – the meaning of politics and the value of democracy for them?
In ImagiDem, we study visual participation both online and offline. We analyse images, stories and memes posted on social media, and follow young people's visual ways of participation as part of their everyday actions. The project combines visual ethnography with computational big data minining and analysis, and deploys this combination to a comparative research setting in four European countries: Finland, France, Germany, and Portugal.
The ImagiDem project has begun by an intensive period of both data collection and methodological development. During its first half, the project members have conducted ethnographic fieldwork in three countries (Finland, France and Portugal), and on altogether six fields – thematic objects of visual participation and politics – so far.
Along with data production, a major focus of fieldwork was the development of a new method – snap along ethnography – providing means to study visual political action simultaneously offline and online. All of the project’s ethnographers have contributed to the testing and developing of the method, which has now resulted in an article manuscript under review. This methodological development was one of the main aims of the project, and is proven to be accomplished very successfully.
Furthermore, the project team has worked to develop new computational tools to study online images. ImagiDem’s unique approach to big data is to create an ethnographically based neural network training set that can identify and classify large sets of online images automatically.
In our ethnographic subprojects, we have investigated how young activists use images and image-related practices in their activism: how they politicise visually and what is the significance of images and their taking for their activism. In Portugal, the project has followed climate activists and the housing movement. In France, we have investigated the yellow vests, and the climate movement. In Finland, the project’s focus has been the climate movement and experience-based activists (esp. homelessness and mental health activists). Fieldwork has been conducted simultaneously online and offline, offering a unique 360-degree perspective on the youth’s visual political action. In addition to participant observations, altogether 100 young activists have been interviewed by the team across three countries. The interviews have included co-analysing the research participants’ images. The ethnographic data produced are extremely rich and the analyses thereof have already resulted in 5 peer-reviewed articles, two submitted manuscripts and several manuscripts in preparation.
The computational subproject has achieved a point in which we are now able to implement a system that automatically recognizes digital images according to a categorization made by the team’s social scientists based on their ethnographic fieldwork. The work started with the creation of a training set of 20 000 images downloaded from Instagram based on relevant hashtags identified via ethnographic fieldwork. By using a feature extraction algorithm on this training set, we then trained a neural network to identify different visual types of politicization and political participation. By way of labeling and re-labeling based on the ethnographic knowledge acquired, we taught the algorithm to classify the images into seven categories of visual political action, identified equally on the respective ethnographic fieldwork. The algorithm has now been tested on 200 000 images on climate activism, and it is ready to be employed for comparative analysis across fields, countries and time periods.
Along with data production, a major focus of fieldwork was the development of a new method – snap along ethnography – providing means to study visual political action simultaneously offline and online. All of the project’s ethnographers have contributed to the testing and developing of the method, which has now resulted in an article manuscript under review. This methodological development was one of the main aims of the project, and is proven to be accomplished very successfully.
Furthermore, the project team has worked to develop new computational tools to study online images. ImagiDem’s unique approach to big data is to create an ethnographically based neural network training set that can identify and classify large sets of online images automatically.
In our ethnographic subprojects, we have investigated how young activists use images and image-related practices in their activism: how they politicise visually and what is the significance of images and their taking for their activism. In Portugal, the project has followed climate activists and the housing movement. In France, we have investigated the yellow vests, and the climate movement. In Finland, the project’s focus has been the climate movement and experience-based activists (esp. homelessness and mental health activists). Fieldwork has been conducted simultaneously online and offline, offering a unique 360-degree perspective on the youth’s visual political action. In addition to participant observations, altogether 100 young activists have been interviewed by the team across three countries. The interviews have included co-analysing the research participants’ images. The ethnographic data produced are extremely rich and the analyses thereof have already resulted in 5 peer-reviewed articles, two submitted manuscripts and several manuscripts in preparation.
The computational subproject has achieved a point in which we are now able to implement a system that automatically recognizes digital images according to a categorization made by the team’s social scientists based on their ethnographic fieldwork. The work started with the creation of a training set of 20 000 images downloaded from Instagram based on relevant hashtags identified via ethnographic fieldwork. By using a feature extraction algorithm on this training set, we then trained a neural network to identify different visual types of politicization and political participation. By way of labeling and re-labeling based on the ethnographic knowledge acquired, we taught the algorithm to classify the images into seven categories of visual political action, identified equally on the respective ethnographic fieldwork. The algorithm has now been tested on 200 000 images on climate activism, and it is ready to be employed for comparative analysis across fields, countries and time periods.
The methodological and theoretical advances in the project push beyond the state of the art in three major ways.
First, the development of the snap-along method opens a new ethnographic approach for studying visual political action in contemporary societies. A method that focuses simultaneously on actions online and offline, and is particularly adapted for the study of visual action, has thus far been lacking in the social scientists’ toolkit. As outlined in the project plan, the method has now been developed in the project, and has been presented at the International Visual Sociology Conference 5.–8.7.2021 Dublin, Ireland (online). The article manuscript describing the method is currently under review in an academic journal.
Secondly, the computational subproject has produced an algorithm that is employable for comparative analysis of large image datasets. This tool will allow for the comparative analysis of visual political action across contexts, themes, and time. Such a tool offers possibilities for analysis that have thus far not been conceivable in the analysis of political action. Towards the end of the project, with the combination of ethnographic work and the novel computational analytical approach, the project will produce new knowledge on how visual political action takes shape, and how it changes and varies across contexts, themes, and time.
Thirdly, the theoretical development on visual politicization and visual forms of engagement is in progress, and is proving to open up a new avenue for conceptualising politicization and for understanding visual political action. Theoretical tools to combine and enhance elements from recent pragmatist theorizing and cultural sociology’s concepts are taking shape in the form of, firstly, a theory of politicization through people’s engagement in the common, and secondly, a new theoretical approach titled new social pragmatism that challenges the dominant understandings of a deliberation based public sphere and its role in democracy. The latter inserts the project’s findings on visual politics and politicization into the pragmatist conceptualisation of public justifications, stakeholder interests and emotional political attachments. Both these theoretical developments and currently taking the form of first publications, and will by the end of the project develop into full-fledged theoretical advances.
First, the development of the snap-along method opens a new ethnographic approach for studying visual political action in contemporary societies. A method that focuses simultaneously on actions online and offline, and is particularly adapted for the study of visual action, has thus far been lacking in the social scientists’ toolkit. As outlined in the project plan, the method has now been developed in the project, and has been presented at the International Visual Sociology Conference 5.–8.7.2021 Dublin, Ireland (online). The article manuscript describing the method is currently under review in an academic journal.
Secondly, the computational subproject has produced an algorithm that is employable for comparative analysis of large image datasets. This tool will allow for the comparative analysis of visual political action across contexts, themes, and time. Such a tool offers possibilities for analysis that have thus far not been conceivable in the analysis of political action. Towards the end of the project, with the combination of ethnographic work and the novel computational analytical approach, the project will produce new knowledge on how visual political action takes shape, and how it changes and varies across contexts, themes, and time.
Thirdly, the theoretical development on visual politicization and visual forms of engagement is in progress, and is proving to open up a new avenue for conceptualising politicization and for understanding visual political action. Theoretical tools to combine and enhance elements from recent pragmatist theorizing and cultural sociology’s concepts are taking shape in the form of, firstly, a theory of politicization through people’s engagement in the common, and secondly, a new theoretical approach titled new social pragmatism that challenges the dominant understandings of a deliberation based public sphere and its role in democracy. The latter inserts the project’s findings on visual politics and politicization into the pragmatist conceptualisation of public justifications, stakeholder interests and emotional political attachments. Both these theoretical developments and currently taking the form of first publications, and will by the end of the project develop into full-fledged theoretical advances.