Data and methods
Our analysis is based on a variety of qualitative research material. In order to examine how self-tracking is presented within the cultural imaginary, we collected textual material from webpages of individuals and organisations that promote and endorse everyday self-tracking. These materials include Quantified Self-related website publications, public blog posts as well as promotional materials of private enterprises that manufacture self-tracking devices. Furthermore, to grasp the experiences related to the therapeutic regimes of action in everyday self-tracking practices, we employed 15 semi-structured interviews conducted with Finnish self-trackers. The interviews were gathered during 2015-2016. Of the interviewees, six were men and nine were women, all of them were employed, studying or both at the time of the interviews. During the last few years, the first author has engaged in diverse self-tracking practices, including an eight-month period of consistently wearing a popular, consumer-grade activity tracking wristband (FitBit Charge HR). This, we feel, has enabled tacit knowledge on the studied field and the ‘workings' of personal self-tracking assemblages in everyday life.
In order to investigate our overarching research questions, we conducted a close reading of our research material through two main themes. First, we looked into how self-tracking technologies and the data they produce are presented as being useful and meaningful for people in self-discovery, self-adjustment and self-improvement. Second, we investigated how the data serve - and how they are narrated as serving - the pursuit of a good life. Our analysis is not focused on
Datafying therapeutic life management 111 human-device relations, and we do not try to individuate patterns of use for the deployment of self-tracking devices. Instead, we study self-tracking as a practice and technology of the self (Foucault, 1988). This means that we have analysed how self-tracking as a regime of action springs from, mediates and shapes one’s relation to the self and how the self becomes examined, understood and enacted through self-tracking. In addition, the concept of assemblage works as a methodological stance that enables us to focus on constant change rather than stability. This is to say that our analysis is sensitive to the premise that things - such as ‘the self, the ‘therapeutic’ or the practice of ‘self-tracking' - are constantly put together from diverse elements and are in a constant flux rather than fixed in place through notions of static ‘essences’ or ‘cultures’. In relation to self-tracking as a proliferating practice of life-management, then, our aim is to show that there is no unequivocal way in which self-tracking ‘is’ a therapeutic practice of self-care. Rather, self-tracking ‘becomes’ therapeutic relationally, that is, in relation to the sociotechnical and political context in which it is practised.
The fragmentary holism of self-tracking
Self-tracking technologies are often implemented in everyday life and social imaginaries in manners that resonate with the therapeutic ethos of self-discovery and the pursuit of a self that is somehow ‘whole’. Perhaps already due to the term se/f-tracking, the practice often becomes associated with the management of an undeniable uniqueness and wholeness of the person that an idea of the ‘self stands for. Yet, through general calls for ‘self-knowledge through numbers’, as the Quantified Self movement’s popular slogan goes, people are encouraged to assemble self-knowledge through a wide variety of quantified data (Lupton 2016; Berg 2017). Different instances of and devices for self-tracking measure different and often very limited aspects of personal being - such as a step count or a heart rate. Therefore, self-tracking attracts very specific modes of knowing about personal existence. In being specific in this way, the activity already builds on a quite special notion of selfhood that steers away from looking at ‘unique’ life conditions and settings. In this section, we examine how self-tracking is framed as a therapeutic practice of making sense of the self, and how it is presented as a means for self-improvement. Through this analysis, we shed light on a contradictory tension internal to the dynamics of self-tracking, that we call fragmentary holism.
Our research materials show a plethora of ways by which technology developers narrate self-tracking devices’ holistic capacities. Take, for example, Polar Electro Ltd., the manufacturer of popular fitness tracking wristbands that gather data on daily step counts and patterns of sleep. The company claims that monitoring by the wristband helps the person to become physically more active which, in turn, reduces health risks, increases personal well-being and improves general vitality. In the developer’s promotional words, the device provides a ‘complete’ and ‘truly holistic’ picture of daily activity and ‘highlights the importance of every movement’.2 Similarly, the Finnish wellness ring manufacturer Oura claims that ‘[w]ith Oura, you learn your optimal times to move, eat and take a break toget that restorative sleep’.3 The ring collects data on, for example, heart rates and nightly heart rate variability to inform users of their sleep quality. The developers call this a ‘holistic method [...] built on years of experience in human performance and the study of circadian rhythms of the body’. Taking another example, HeartMath Inc. is a technology developer that focuses on stress reduction via the measurement of heart rate patterns. The company promotes devices such as the ‘inner balance’ sensor, and a method for constant monitoring of the changes of the heartbeat by claiming that such tracking leads one towards a state of coherence as long as one ‘stays with it’.4 It further claims that measuring heart rate helps the person to ‘incorporate the heart's intelligence into their day-to-day experience of life’ and to ‘become the best version [of oneself] more often’. As anthropologist Natascha Dow Schull (2016) noted in her ethnographic study of health tech exhibitions in the US, the self-tracking industry’s marketing constantly implies that the choices we make (in terms of attaining activity, good sleep, relaxation, etc.) reflect something crucial about who we really are.
In the above examples we see an adaptation of the therapeutic language of holistic self-understanding brought into the domain of digital self-tracking. The developers’ claims often imply that the person may become something else as a whole through permanent tracking. For example, the individual is supposed to increase personal ‘general vitality’ and become ‘physically more active’ through step-focused activity tracking, or find ‘coherence’ and ‘inner balance’ via heartbeat data. Yet the tension between a holistic approach to personal life (referring to transformation of the individual as a whole) and ‘fragmentary holism’ (referring to linear optimisation of nuanced functions of the body) becomes evident in the marketing language. The developers paradoxically encourage a better holistic version of oneself (e.g. a more ‘vital’ or ‘balanced’ self) by focusing on an algorithmically predefined functionality (such as a step count or heart rate) or necessarily limited combinations of such functionalities.5 In addition, by encouraging people to become an improved version of themselves more often - for instance daily - the self-tracking imaginaries refer to an ongoing struggle of transformation. Thus, Polar, Oura and HeartMath end up presenting self-improvement as a linear and ongoing process (see Bode & Kristensen, 2015) that is about constant monitoring and constant potentiality’ rather than about actuality.
Fragmentary holism highlights the idea that the ‘self is brought forth as an assemblage - disassembled and put back together through data - in self-tracking practices. First, the self consists of limited ‘functions’ or ‘parameters’ such as heartbeats or body movements, which are fragments of living that can be measured - and importantly, separated and combined - by means of self-tracking. In the case of an activity tracker, for example, a seemingly general picture of ‘healthiness’ or ‘vitality’ can be assembled by combining measurements on, for example, movement, heart rate and sleep quality. Second, through longitudinal measurement of any specific fragment or vital function, the self is enacted as a data assemblage. Several scholars point out the relational character of tracking as a practice that requires assembly work on separate yet entangled data points or data nodes (Ruckenstein, 2014; Day & Lury, 2016; Schull, 2016). This means
Datafying therapeutic life management 113 that in order to make sense of and indicate progress or regress, any individual measurement (e.g. a step count) needs to be related to other similar measurements at different points in time.
We can illustrate the idea of the self as a data assemblage through the typical features of self-tracking software. Most self-tracking software visualises self-related data as graphs and charts that can be employed for self-development purposes. For example, the Polar Loop wristband gathers individual steps into longitudinal (daily and weekly) indicators of activity. In this way, the individual is disassembled into individual units of movement by the device’s three-dimensional accelerometer, and re-assembled by the software. For example, depending on the device and software, a daily step assemblage on the application screen can take the form of a loop that gradually closes as one gathers steps and nears the daily goal, or a bar that fills up accordingly. A weekly activity assemblage may be visually presented e.g. as a bar chart. Similar logic pertains to measurements of heart rate, sleep, etc.; individual beats or movements are assembled into daily/nightly stats, which are then further assembled into longitudinal graphs. It is also typical that a simple number indicates the tracked self in a moral sense. For example, a self-tracking application may make complimentary claims about the person based on their total number of daily or weekly steps, or offer an ‘efficiency percentage’ for daily/weekly activity or sleep. Achieving high numbers or hitting 100 percent are then often virtually rewarded with animations, colour codes, trophies, etc. Here individual measurements or measurement combinations are algorithmically assembled into an evaluation. It is typical that such ‘holistic’ indicators of the self can again be disassembled into smaller-scale information in order to get more nuanced data on the specific function, for instance, to see the distribution of steps on a daily or hourly basis. However, in everyday use, to see the ‘importance of every movement’, as Polar mentions as a goal, is to ultimately trust the algorithm that assembles a personal evaluation. However, completion of a single evaluative checkpoint (e.g. a single day) is never really a completion at all. For example, in the Polar Loop activity tracking software, the person is implicitly encouraged to reach 100% activity (i.e. enough movement) daily. However, the weekly visualisation of one’s activity shows a ‘daily goal completion average’, which groups together daily evaluations and offers a different number, the average.
Considering the self as a data assemblage also means that individual measurements of different functions could or should also ideally be related to each other. For example, according to Polar, when combined with a heart rate monitor, the activity wristband can become an even more precise indicator of ‘activity’, as it can then offer more accurate quantification-based insight on the intensity of activity conducted, calories burned, etc. It can be argued that the more elements the assemblage has, the more ‘holistic’ the image of the self it paints. However, the more elements there are, the more data there are on separate functions that can, or need to, be tied together and related to other measurement data.
The point here is that as the tracked self is enacted as a data assemblage, selftracking as a way of working upon the self always feeds back to the user through its limitations. Data could always be more complete and more saturated throughrelations, references and repetition. Therefore, the practice is in principle opposed to the idea of a ‘holistic’ understanding of the self. Of course, we do not deny that self-trackers often pay attention to qualitative data - for example, their experiences and sensations - in order to make deeper sense of the quantitative data (see e.g. Ruckenstein, 2014; Lupton, 2016). However, while we grant that people may well experience the gaining of insights into their lives by contextualising the data, we think that holistic humanism and reflexive hermeneutics of the self often tend to be eclipsed or even be made redundant by the technical affordances of self-tracking: the precision of measurement of a specific function or activity at a specific point in time. Polar, for example, directly presents the process of ‘becoming a more active person’ as a project in which the person becomes conscious of her/his step-based physical movement within a day in specific ways and adjusts her/his behaviour to modify these specific modes of action. Such awareness and a trajectory of personal action are induced by the measurement because physical movement (or lack of it) is what the devices and algorithms recognise, react to and respond to by producing data that then provide feedback for the user.
To sum up, from the viewpoint of assemblages of control at work in selftracking, individuals become increasingly conceived of as compounds of functions, parameters, capacities and resources to be regulated and operated upon. Such control as a mode of governing people and their lives sits uneasily with the holism that nevertheless surfaces as an emic emphasis in the imaginaries of the capacities of self-tracking. For holistic therapeutic practices, the individual person and her/his ‘unique’ life, experiences and aspirations are the focus of practices to govern and modify people, whereas control and dividualisation reach beyond and below the personal by means of the collection and aggregation of masses of data. We claim that while many therapeutic practices are concerned with encouraging people to discover their unique modes of action for self-improvement, digital selftracking today becomes functional by focusing on self-control and by improving modes of action already predetermined in the devices’ code. In what follows, we will take a closer look at everyday experiences of self-tracking and discuss these experiences as reflective of anticipatory regimes of action.