Guest authors: Daan Doeleman, Donovan van der Haak, and Giovanni Prins
Does (mental) health tracking technology help people become more autonomous? Our answer is that it may, but whether and how it does so depends strongly on what is understood under ‘autonomy’.
In this discussion, we claim that self-tracking technologies that are designed through the lens of dominant individualist conceptions of autonomy risk uncritically and unnoticedly nudging its users and the society in which they are embedded into a self-understanding that reinforces this individualistic conception, and may therefore actually threaten other (non-individualist) conceptualisations of autonomy that fit better with ideals of care and mutual dependence between people.
By connecting debates about conceptions of autonomy to the (post-)Foucauldian literature on subject-formation and responsibilisation, we hope to suggest how design for autonomy, particularly in mental health tracking technologies, stands to gain from relational conceptions of autonomy.
Self-Tracking Apps
In the 21st century, evaluating one’s own (mental) health has become increasingly common. Combined with the availability of (relatively) cheap wearables, widespread use of smartphones and other means to track health data, this has led many, at least in affluent parts of the world, to track and thereby attempt to improve their (mental) health. The Apple Health app, for example, includes a function to record one’s emotions and moods. Zepp offers a ‘stress detector’ that bases this on heart rate data through its smart watches.
The underlying thought is that collecting this data allows one to take action, perhaps by building new habits such as eating healthier, sleeping better or going for a walk (patterns that can themselves be tracked again), or even by seeking out professional help. In the words of early self-tracking advocate Gary Wolf, who coined the term ‘Quantified Self’: we can use data to “reflect, learn, remember and improve… If we want to become more effective in the world, we have to know ourselves better.”
Social & Societal Dimensions
Although this may sound promising, such self-tracking approaches to (mental) health can be thought to exemplify what Van Grunsven (2024) criticises in the “Ableist Cartesian Sociotechnical Imagination” (1): a way of thinking about technology that is strongly individualist and overlooks embodied and relational aspects of human being and even presents a ‘threat to human flourishing’ (Ibid.).
She articulates her criticism mainly in the context of bodily health care technologies and the design imagination shaping them. We argue that a similar Cartesian – or as we will claim, neoliberal – individualist notion of the self-sufficient agent underlies – and is reproduced in – present day thinking about (mental) health more generally and the self-tracking technologies devised to support and strengthen it.
Users can of course use self-monitoring technologies to become more psychologically resilient by gaining self-knowledge, thereby becoming more autonomous. At the same time, these technologies may lead them away from alternative practices of taking responsibility for oneself and others, and from interpersonal ways of coming to self-knowledge about our own health (which also includes assessments of bodily aspects of how we are doing) by reinforcing a particular, more individualised notion of autonomy. In the following, we would like to demonstrate how (post-)Foucaultian literature can help show this.
Disciplinary Power and the Individual
In Discipline and Punish, Foucault (2020) gives us an account of how the individual is produced through a complex process of discipline and individualisation. Famously, Foucault argues that modern societies have become characterised by “relations of discipline” (Ibid., 208). Power, instead of being concentrated in “the body of the king” (Ibid.) had slowly become decentralised and dispersed throughout many facets of society.
This decentralisation and dissemination of power meant that the effects of power were brought “to the most minute and distant elements” (Foucault 2020, 216). In other words, power was exercised not by a single authority, but throughout all layers of society. Foucault’s most famous example is the prison, which exercised power through the faceless architecture of the panopticon, inducing in the prisoner a “permanent visibility that assures the automatic functioning of power” (Ibid., 201). But, the disciplinary apparatus was not limited to the prison. Indeed, it was dispersed throughout all layers of society – e.g., schools, the military, hospitals (Ibid., 146-147, 170-172, 193).
A Foucaultian analysis of discipline shows how the way disciplinary power is exercised makes individuals (Foucault 2020, 170). By showing how concrete historical processes and practices have led to the creation of particular subjectivities and social categories, Foucault illustrates how the subject is not an essence that pre-exists the social world, but is rather historically contingent. For example, ‘the delinquent’ seems to be a natural and pre-given entity with certain essential characteristics. However, Foucault argues, this category is a historical product, and the figure of the delinquent is produced in and through the disciplinary apparatus (Ibid., 253).
In this process of individualisation, Foucault (2020) explains, the individual internalises the disciplinary power from which it is wrought, “since we are part of its mechanism” (217). That is to say, the individual internalises the codes and norms that produce it, and it subsequently governs itself in accordance with those norms. In this internalisation of power, then, their (self-)conception and status as individuals is constructed. This becomes all the more clear if we return to the case of self-monitoring technologies.
Self-Monitoring and Discipline: Normalisation and Responsibilisation
While Foucault’s account of discipline is nothing new, technological advancement potentially exacerbates - or at the very least makes use of - disciplinary mechanisms. Self-monitoring is a particularly pertinent example of this. Self-tracking apps such as the Apple Health app or Zepp allow the individual to continuously check on their wellbeing, and improve it in the relevant aspects. As mentioned above, it allows analyses of one’s emotions and moods, but also quality of sleep and other relevant data on one’s health. By monitoring these, users are constantly exposed to and aware of this data, as well as the possible space for improvement.
The disciplinary power that self-monitoring technologies exercise on individuals functions through such ways of generating information. People are made to behave in the way they are ‘supposed to’ through a process of being made visible and being evaluated in terms of created social roles with accompanying norms. Other than in Foucault's prime example of the prison, however, the behaviour that is stimulated by self-tracking apps is - at least prima facie - self-chosen.
One may therefore object at this point and wonder why this is a problem. Surely, this increase in self-knowledge and consequent enhancement of individual autonomy to reach personal goals more effectively is precisely the aim of these technologies. While this is certainly true, we argue that self-monitored mental health are symptomatic of – and thereby cultivate - the neoliberalisation of care practices.
Technologies such as the ones mentioned above rely on and subsequently reproduce a notion of subjectivity that is characterised by a profound sense of individuality. Moreover, this individuality has the semblance of a necessary and natural aspect of human beings, rather than as the historically contingent norm that it is. Such an individual approach to responsibility may be fitting in some instances, but not always and for everybody. Put differently, there is a risk of responsibility shift. That is, people using these technologies will increasingly be considered – and held – responsible by both themselves and others for their (mental) health. The point then is that the mechanisms that can empower users, and perhaps even expand their sense of autonomy, also mould them in accordance with a contingent norm that emphasises efficiency, self-sufficiency, and absolute individuality. While these technologies thus may enhance individual user autonomy, they also responsibilise users (Butler 2015, 144).
This process of responsibilisation carries several risks. First of all, these individualistic ideas about responsibility and practices of holding responsible may spill over to non-users as well, which is even more problematic from the perspective of autonomy: they may not want to adhere to the norms that users try to adhere to, or they may not be able to self-manage their health in ways suggested by the app. It seems clear that it would be inappropriate to hold such people responsible for their (mental) health in the same way.
But perhaps the most important risk to us is that the individualistic perspective on responsibility misses the intrinsic sociality and interdependence of human beings. The responsibilisation that is instilled and naturalised through neoliberal discourse, and exacerbated via these self-monitoring technologies, effectively erodes institutional care practices and other support networks (Brown 2021, 29), by inappropriately attributing responsibility to the individual. The individualisation of (mental) health via self-monitoring runs the potential risk of turning people into “enterprising subjects” (Chiang, Achaa, and Ball 2024, 3) or “portfolio[s] of self-investment” (Brown 2021, 38), always obligated to enhance and improve oneself.
This ongoing attribution of responsibility to the individual level can be thought to have several adverse effects: it threatens to suppress the social responsibilities to care for others and of others to care for us, it shapes ways of coming to self-knowledge about (mental) health, and this particular way of making visible mental health also suggests very individualistic approaches to improving it. In short, it individualises our ideals of responsibility, the epistemics of (mental) health and the ways of addressing (mental) health issues.
Furthermore, this individualisation of responsibility may obscure the structural obstacles to (mental) health and wellbeing. Instead of recognising that “many forms of vulnerability are caused or exacerbated by social and political structures” (Mackenzie 2014b, 33), the neoliberal ideal makes the individual wholly responsible for one’s own wellbeing.
In sum, while technology-induced autonomy boosting may be very empowering for some users of such technologies, it also risks nudging users increasingly into a position in which they are held responsible for their own (mental) health. In addition, there is a danger of this notion of responsibility spilling over to non-users of these technologies, because such technologies and the type of subject they produce reproduce societal norms about self-governing and personal responsibility for (mental) health. Such notions may undermine ideals of care and the recognition of mutual dependence, as well as more socially oriented ways of coming to self-knowledge. What we might need, then, is an approach to technology design that is more sensitive to our relationality and interdependence.
Design for Relational Autonomy?
Design for Values (Van de Poel, 2013) is a methodological approach that integrates moral and societal values into technological design. One key example is design for human autonomy, which aims to create technologies that support users’ capacity for self-determination and ability to critically reflect on technology/specific technologies. Design for autonomy approaches recognise that technology shapes human action and experience, and strive to prevent manipulative or coercive effects. By embedding autonomy-enhancing features, such as transparency and user control, designers (for autonomy) seek to align technologies with ethical ideals of human agency, ensuring that users can act autonomously and critically according to their own informed choices.
Design for autonomy approaches are therefore of great value. Self-monitoring technologies are a particularly interesting example, as these are not merely designed with autonomy as a by-product, but as one of its primary objectives. In other words, self-monitoring apps are made for self-sufficiency, and are thus exemplary for technologies with autonomy in mind. The perspectives presented above show, however, that autonomy-enhancing technologies such as these may also discipline users in accordance with a neoliberal socio-technical perspective, with the aforementioned risks as a result.
To some extent, the threats of autonomy-enhancing technologies can be well-accounted for within design for value approaches, for instance by designing for different values, or by broadening one’s understanding of autonomy. To do this, however, value-by-design practitioners need to be conscious of the particular subjectivities self-monitoring technologies produce, the underlying ideals that are reproduced by the technologies, and the responsibility attributions that they result in. A Foucaultian analysis of individualisation as suggested above provides the theoretical means to localise these additional risks. Other theoretical means, including more relational accounts of autonomy, can help address some of these risks.
Again, autonomy-enhancing technologies, even when entrenching neo-liberal norms, may still be of great value to individuals, and we certainly do not argue against the use of self-monitoring apps altogether. Arguably, a lot of people stand to benefit from gaining more information, and thereby control over their (mental) health. Individualistically perceived ways of improving one’s (mental) health may be very effective in some cases.
The problem stems from how neoliberal norms of individuality, responsibility, and wellbeing are naturalised, i.e. made to seem necessary, essential, or at least unquestioned. Technological artefacts, even when designed for autonomy, may reinforce neoliberal processes of individualisation. The ostensible necessity of these processes of individualisation, however, may conceal other values and ways of thinking about autonomy. Therefore, a comprehensive approach to understanding the relation between autonomy and technology should involve both designing technological artefacts for autonomy and understanding and evaluating how technological artefacts themselves induce particular norms and values, opening up new possibilities to reexamine values such as autonomy.
For instance, from a relational perspective, we may understand autonomy from the perspective of mutual togetherness, interdependence, and care. Relational approaches to technology design for autonomy emphasise the interdependence between users, technologies, and social contexts (MacKenzie 2014a; Anderson 2022). By adopting a relational approach to enrich our understanding of human autonomy, we can both design technological artefacts with self-determination in mind, as well as combat larger processes of individualisation and responsibilisation.
References
Anderson, Joel. 2022. “Scaffolding and Autonomy.” In The Routledge Handbook of Autonomy, edited by Ben Colburn, 158–66. London: Routledge.
Butler, Judith. 2015. Notes Towards a Performative Theory of Assembly. Cambridge: Harvard University Press.
Brown, Wendy. 2019. In the Ruins of Neoliberalism: The Rise of Antidemocratic Politics in the West. New York: Colombia University Press.
Chiang, Tien-Hui., Achaa, Lydia Osarfo., and Stephen J Ball. 2024. “Activating self-monitoring through the discourse of fear and hope: The subjectivation of enterprising teachers.” International Journal of Educational Research 125, 102324, https://doi.org/10.1016/j.ijer.2024.102324
Foucault, Michel. 2020. Discipline and Punish: The Birth of the Prison. London: Penguin Books.
Grunsven, Janna B. 2024. "Disabled Body-Minds in Hostile Environments: Disrupting an Ableist Cartesian Sociotechnical Imagination with Enactive Embodied Cognition and Critical Disability Studies." Topoi: An International Review of Philosophy. https://doi.org/10.1007/s11245-024-10080-5.
Mackenzie, Catriona. 2014a. “Three Dimensions of Autonomy: A Relational Analysis.” In Autonomy, Oppression, and Gender, 15–41. Oxford University Press.
Mackenzie, Catriona. 2014b. “The Importance of Relational Autonomy and Capabilities for
an Ethics of Vulnerability” In Vulnerability New Essays in Ethics and Feminist Philosophy, edited by Catriona Mackenzie,Wendy Rogers, and Susan Dodds. Oxford University Press.