Personal data are political . A feminist view on privacy and big data

The second-wave feminist critique of privacy defies the liberal opposition between the public-political and the private-personal. Feminist thinkers such as Hanisch, Young or Fraser note that, according to this liberal conception, public institutions often keep asymmetric power relations between private agents away from political discussion and action. The resulting subordination of some agents to others tends, therefore, to be naturalised and redefined as a «personal problem». Drawing on these contributions, this article reviews the social and political implications of big data exploitation and questions whether personal data protection must remain a matter of «privacy self-management». It aims to show that feminist political theory can decidedly help to identify and tackle the root causes of what I call «data domination».


INTRODUCTION
Daily life is increasingly mediated by new information and communication technologies. The use of digital and online services is necessary to live a normal life in Western countries. It is unthinkable, for example, to complete a university degree or to perform certain jobs without using a computer, a mobile device or the Google search engine. Online social networks permeate social relations. Access to basic services such as health, banking or urban transport has been digitised. All this has led to a datafication of everyday analogue and digital life (Baruh & Popescu, 2015). The velocity, variety and volume of data generation have increased dramatically since the late 1990s, up to the point of submerging society in a «big data ecosystem» (Cox & Ellsworth, 1997;boyd & Crawford, 2011).
Our normal actions and interactions generate great amounts of personal data that spread further, faster and more broadly than ever before throughout a space controlled by private interests (Zuboff, 2019). Data can form a complex profile of our attributes, interests and preferences, but also about our environment and the persons who are part of it. In this sense, even if you do not use the services or products provided by a certain corporation, it can get information about you through the data other people generate by using those services or products, or thanks to your use of the products or services provided by related companies. By way of example, you can delete your Facebook account (or you can never have had one) and Facebook can still having information about you: this is the case of «shadow profiles» (Sarigol, García & Schweitzer, 2014). Furthermore, data disclosing is often automatic and unconscious. The case of metadata exemplifies it particularly. Due to all of this, it is very difficult, if not impossible, to actually prevent one's own life from being recorded in data (Suárez-Gonzalo, 2017).
As predicted by Lesk (2001), the velocity, variety and volume of big datasets undermine humans' capacity to make data «useful». Therefore, big data technologies draw on a new logic of data collection and analysis, which allows inferring latent information from datasets that isolated data do not reveal. At present, these techniques are necessary to decide even which of all these data deserve human attention. Furthermore, automatic mechanisms replace important human actions, as decision-making or the analysis and classification of places, objects, ideas, ways of expression and behaviours. This new «algorithmic culture» (Hallinan & Striphas, 2016) involves what, in terms of Federici (2004), would be a mechanisation of big data production and reproduction. On the one side, every human activity is systematically converted in data. On the other, humans are conceived as data production machines. But, does this technological development primarily contribute to social good and equality? Or does it privilege the interests of a few?
Over the past few decades, a new business model focused on data exploitation has emerged. Data business is an opaque sector, characterised by a high business concentration that has monopolised big data turning them into a gold mine. The so-called GAFAM corporations (Google, Apple, Facebook, Amazon and Microsoft) control the data market. They gather the most of the data that billions of people disclose every day (Tufekci, 2017). As a result, both the economic value and the value for social control of the data collected by GAFAM grew exponentially. Indeed, one of their main achievements has been to start addressing data as interchangeable goods, opening the markets boundaries into them, and so paving the path for extracting profit. A practice that frames in what several authors (see Harvey, 2004;Fraser, 2016;or Arruzza, Bhattacharya & Fraser, 2019) define as a privatising, financialised and predatory form of capitalism that appropriates, rather than generates, wealth and income. Harvey (2004) uses the expression «accumulation by dispossession» to designate the mechanisms whereby this renewed expression of capitalism gets richer.
Although there are symptoms of public dissatisfaction with this situation, individuals' data-disclosing behaviours do not reflect it: people affirm to be concerned about the lack of protection of their personal data, but carry on disclosing their information in a seemingly carefree way, especially on social networks. This dissonance is known as the «privacy paradox» (Barnes, 2006). How should we interpret this apparent contradiction? Are people concerned about how big data exploitation can affect their fundamental rights to privacy and personal data protection? Are their behaviours due to ignorance, conformity or irresponsibility? Or is it perhaps a problem of impossibility to act according to the own concerns?
My point is that the privacy paradox is, primarily, a consequence of the inequality of power between citizens and those who exploit their data. While companies and corporations establish the rules for new communication and information practices, impose the data disclosing conditions through their «privacy policies», control the data flows and benefit from them; citizens generally do not have basic information about what happens with their data, nor the means to handle the situation or to escape it. I define this asymmetric power relation as «data domination».
Following Pettit (1997;, an agent (that can be an individual, a group of people or a collective or a corporation) dominates another (often individual) if he or she has a power to influence the life or decisions of the other agent that this latter does not itself control. That is what Pettit (2012) calls a power of «uncontrolled interference» (also referred to as «arbitrary interference» in Pettit's earlier works). This statement is still valid if the dominating agent is never going to make «effective» her or his power, or if the dominated agent is not aware of her or his vulnerability. In that line, Fraser (2012) holds that subordination is linked to the impossibility of detecting injustice. According to her, in a morally fair social order, everyone must have access to the same means and resources. To dominate another involves frustrating her or his possibilities to notice what is fair and unfair and/or to claim and act for justice. Thus, the lack of explicit protest should not be interpreted as an expression of justice. Combating injustice requires discursive resources and interpretative schemas that allow its identification and denounce.
It should be noted that, the monopolistic power of data markets would not be possible without the black boxed condition of the algorithmic culture. Following Pasquale (2015), corporations and governments foster an intended opacity through real and legal secrecy and obfuscation. He uses the term «agnotology» to describe the «structural production of ignorance, its diverse causes and conformations, whether brought about by neglect, forgetfulness, myopia, extinction, secrecy, or suppression» (2015: 2). This opacity reinforces people's technological disempowerment: they have the power to use technology, but not to understand it.
Facing this situation, the European legal framework for the fundamental right to personal data protection (hereinafter «PDP») -recognised by both the General Data Protection Regulation 2016/679 (hereinafter «GDPR») and the Charter of Fundamental Rights of the European Union, Art. 8 (2016/C 202/02)-is based on the liberal-hegemonic conception of privacy (Baruh & Popescu, 2015;Richardson, 2016). GDPR has been widely recognised as one of the most ambitious measures to fight personal data's vulnerability. It entails meaningful progresses in relation to the former Directive of 1995. Nonetheless, I argue in the following pages that, as feminism has shown, the liberal ideal of privacy is unable to protect people from domination. The aim of this article is to show, in light of the second wave feminist critique of privacy, that the conceptual framework of GDPR makes it unable for citizens to protect themselves from data domination. In following this objective, the text is in two main parts: The first one focuses on the liberal-hegemonic conception of privacy that underpins GDRP. The second one explains the second-wave feminist critique of privacy and what can we learn from it in relation to data domination. Drawing on this reflection, I stand up for rethinking privacy and so personal data protection from a feminist point of view.

LIBERAL PRIVACY: AN INTERFERENCE-PROOF BUNKER
Traditionally, the private sphere has been opposed to the public one. On the one side, the public relates to the community and its concerns. Politics literally means the things concerning the polis. Then, preserving the public has been historically a responsibility of politics, which, at least since the Modern Age, has to do with the organisation of the government and the state. On the other side, the private refers to the personal, the domestic and property and identifies with the secret, the isolated and the hidden. Within this logic, the private has been traditionally excluded from public discussion and action, and understood as a matter of self-regulation (Mill, 1992). The limitations of this opposition between the public and the private spheres are discussed throughout the following sections.
In 1890, Warren and Brandeis famously defined the right to privacy as an extension of the right to «be let alone». A right «against the world», based on the principles of inviolability of personality or immunity of the person, and related to the protection of those aspects of life whose invasion supposes an injury to feelings. Including thoughts, emotions, unpublished productions of the intellect, sayings, acts, personal relations, domestic issues, etc. According to Warren and Brandeis, the right to privacy finishes «upon the publication of the facts by the individual, or with his consent» (1890: 8). The law, so, must have the purpose of protecting each individual from the unwarranted invasion in those private aspects with which the community has no legitimate concern. This negative conception of privacy (as an «inviolable» sphere) is linked to liberal theory, which defines freedom as non-interference (see Berlin, 1969: 196 andHobbes, 1994: 187). From the liberal perspective, a person is free if there are no interferences that modify her or his course of action. Thus, freedom consists in doing what one wants to do, without being forced to do so. In that sense, capacity is not a condition for freedom: not having the means or the ability to do something does not mean that one is not free to do it (Pettit, 1997).
If we understand freedom as the absence of interferences between each other's actions, to be free is only possible in isolation. Because of that, several authors affirm that liberalism understands and analyses society as a sum of atomised individuals (see Bertomeu & Domènech, 2005). Hence, liberal freedom is, by definition, in conflict with social life. Nonetheless, liberal theory understands that, to guarantee that one's freedom does not harm that of others, and thus ensure other values such as security or property, living together requires the imposition of certain limits to individual freedom (Mill, 1992;Hobbes, 1994). This relates to Berlin's (1969) «value pluralism», according to which the main moral values are either equally important or incommensurable. This is a form of moral relativism that implies that there is no rational way to resolve conflicts between values. Consequently, the liberal state must refrain from promoting a certain ideal of good living. The problems of this vision have been stressed by Dworkin (2011).
In light of this theory, privacy is the individual sphere away from many of the «necessary» restrictions that life in common imposes on individuals' will when they participate in the public sphere. Liberal privacy is, let us say, a bunker of «pure» freedom, proof against interferences. The right to privacy is, then, a right to enjoy a personal sphere exclusively governed by the individual will.
Contemporary liberalism has criticised some aspects of this traditional way of understanding how privacy must be protected. Some of them, as Etzioni (1999) denounces that privacy has a potential to conceal actions that may threaten public order and, particularly, security. He therefore argues for reinforcing the «limits» of privacy, or, in other words, the need to restrict the «free will» in the private sphere. Solove (2008), for his part, holds that privacy is a «contextual» value that should be defined on a case-by-case individual basis, in relation to the harm (or the benefit) caused to the individual from its violation. It is up to each individual to determine which aspects of the own life he or she keeps in private or shares in public, depending on what he or she wins or loses by doing so. Regarding the protection of privacy, Solove (2013) proposes a form of «libertarian paternalism». Taking into account the fact that people's decisions or actions to protect their private affairs are often inappropriate, he advocates replacing individual decision by an external imposition in those cases in which this may help to keep some aspects of people's lives in private.
As seen, Both Etzioni and Solove claim for the imposition of certain limits to the individual will in the private sphere. By doing so, they nuance the traditional «inviolability» of privacy. Nevertheless, far from seeking alternative terms to redefine its value, they continue defining it in terms of interference.

Liberal privacy in GDPR
GDPR follows the privacy self-management paradigm (Baruh & Popescu, 2015;Solove, 2013). In other words, privacy is the underpinning value of the European PDP regulation, and there is a broad consensus that it should be so. GDPR concretises this paradigm in the notice and choice model (Schwartz, 2013;Solove, 2013) and establishes individual consent as the mechanism individuals have to protect their personal data from massive exploitation. This consent has to be free, informed, specific and revocable, and it must be expressed clearly and unequivocally (see GDPR Arts. 4(11) and 7). Hence, the interested party should notice the individual which specific data is interested in collecting or processing and for which specific purposes. The «data subject» must give her or his consent prior to the gathering and processing of the personal data. Nevertheless, due to the very logic of big data, it is difficult to predict which information could arise from the aggregate processing of a massive dataset. Consequently, GDPR cannot actually guarantee that individual consent is always «specific» and «informed» (Suárez-Gonzalo, 2017).
Furthermore, «free» means in the Regulation that individuals cannot be forced to share their data (see Art. 7), i.e., that consent must be voluntary. In that sense, what GDPR protects is, basically, the inviolability of personal data: the data-privacy of the individual in a liberal sense. On doing so, GDPR places the burden of PDP on individuals and their isolated decisions, turning consent into a sort of private, «voluntary» contract between two unequal parts: the individual and the agent that collect, analyse and use her or his data.
Nevertheless, that «contract» can be unilaterally broken in those cases related to the protection of other rights, considered equal or more important than the right to PDP. In those cases, the cession of personal data is not «free» but «necessary». These include cases in which: its processing is necessary to protect the vital interest of the person; it is justified by legal obligations; for reasons of substantial public interest (including public safety and health); among others. Consent is neither required in case the data subject has disclosed them in a public space; or if the processing is carried out by an organisation not-for-profit with a political, philosophical, religious or trade union aim, provided that the processing relates solely to members or to persons who have regular contact with it and that the personal data are not shared with third parties (Article 9, d).

THE FEMINIST CRITIQUE OF THE DARK SIDE OF PRIVACY
The division of the public and the private spheres in liberal theory and practice finds some of its greatest detractors in the feminist movement. According to Pateman (1989), the public/private dichotomy is what feminist political theory is fundamentally about. The feminist movement noticed the perversity of opposing the public to the private. On the one hand, feminists pointed at a dark side of privacy that has served to hide and legitimise the domination of women. On the other, they showed how personal circumstances are structured by public factors that provoke unequal social orders. The feminist movement stressed that a person cannot be as free and equal in her or his private sphere as those to whom he or she is subordinated in the public one. Furthermore, Pateman says, feminism has devoted great efforts to investigate the contradictions of the liberal state. It brags about stopping at the threshold of the private, while configuring the private by means of laws and policies that subordinate some citizens to others due to their gender, class or race.

The personal is political
The so-called second-wave feminist movement, which arises as part of the radical socialist politics from the 60s to the 80s, was particularly concerned about the problems of the traditional public/private dichotomy. As noted by Agra Romero (2012: 35-36), second-wave feminists «defied the very unquestioned grounds of both traditional and radical politics» and stood up for a «critical review of the conventions about the limits and nature of the political and of politics». They denaturalised the power asymmetries between men and women and redefined the limits of the 'genuinely' political.
The claim «The personal is political», popularised by a paper published by Carol Hanisch in 1970 andrevisited in 2006, delves into an alternative conception of the public and the private. It came out as a reply to Dottie Zellner who criticised the actions of the Pro-Women Line faction, a specific subgroup within the radical Women's Liberation Movement in New York. The main objective of this group was to bring women's «personal problems» into the public arena, by recognizing «the need to fight male supremacy as a movement instead of blaming the individual woman for her oppression» (Hanisch, 2006: 1-2). Concerning this, Hanisch considers that individual struggle is always necessary. Even more because the oppression that affects the personal usually takes place in isolated circumstances, as the home. However, she clarifies that individual fight against oppression is limited for two main reasons: one is that when people are oppressed they act out of necessity, not out of choice; the other is that personal oppressive situations are often structured by public factors. Then, personal liberation is only possible by socially choosing to change the objective conditions of oppression, and this requires wondering about who benefits from this oppression. In other words, what Hanisch explains is that, to solve personal problems, their structural causes should be part of political discussions and actions. Otherwise, these causes become legitimate.
In this respect, Federici's (2004) work is enlightening. It focuses on the historical subordination of women to men and to the «domestic» (or the private). Federici argues that this is not a causal relation, but the result of the transformations of gender relations prompted by early capitalist forms of accumulation that turn the female body into a means for the «re»production of labour.
In line with this, Nancy Fraser (2012) defends the importance of questioning the liberal tendency to redefine structural inequities as personal problems. She expounds the need to investigate the reasons behind those interpretations that attribute people's unfavourable circumstances to their own weaknesses. Instead of looking for individual panaceas to the problems that affect all of us, Fraser advocates examining biases of the public sphere that expose some persons to different moral considerations that impede their equal access to the political voice.
The feminist movement has interpreted Hanisch' claim in various ways. Some have supported the elimination of privacy. Others, its imposition: According to MacKinnon (1989), the right to privacy has failed to protect women's inviolability. The public/private split has left the private domain out of any form of public scrutiny. Because of that, privacy has become closer to a right to impunity of those men who dominated, oppressed and degraded women, than a right to inviolability of the dominated. In MacKinnon's words (1989: 191), «For women, the measure of the intimacy has been the measure of the oppression». In that light, she holds that feminism has to «explode» the private.
For her part, Young (1987) defends that the determination of privacy must remain. She defines privacy as that aspect of the life and activity that any individual has a right to exclude others from, instead of understanding it as what is excluded from the public. In Young's view, «the personal is political» does not imply denying the distinction between public and private spheres in itself, but the social division that their opposition leads to. The unified model of individuality that has been traditionally imposed in public life discriminates those who do not meet the average canons. It has set aside some persons from the public life by condemning some of their personal attributes or actions to remain private. Her proposal elaborates on a «heterogeneous» ideal of public life which recognises and appreciates the differences, instead of excluding them. Thereby, Young considers that, the defence of privacy «has become not merely a matter of keeping the state out of certain affairs, but asking for positive state action to ensure that the activities of nonstate organizations, such as corporations, respect the claims of individuals to privacy» (1987: 74). In this way, she recognises two principles following from Hanisch's claim: «(a) no social institutions or practices should be excluded a priori as being the proper subject for public discussion and expression; and (b) no persons, actions or aspects of a person's life should be forced into privacy.» Contrarily to Young, and similarly to Solove's (2013) libertarian paternalism, Allen (1999) advocates the need to «coerce privacy». A whole lifestyle premised on disclosure ought not to be an option. Therefore, she stands up for externally imposing individuals to keep some aspects of their lives in private. According to Allen, due to contemporary cumulative damages to privacy -particularly generated by new technologies-, people have «an ethical obligation to protect their own privacy». Then, they have to be forced «to have private lives and to live their private lives in private» (1999: 752). Allen (2016) understands the protection of privacy as a ethical responsibility shared by governments, businesses, and individuals. In her view, the role of a good gov-ernment has to be to protect some forms of privacy that individuals do not value.
To sum up, while Young's interpretation of the claim proposes an alternative conception of privacy to the liberal one, MacKinnon and Allen do not actually move away from privacy's «inviolable» conception.

FEMINISM FOR PERSONAL DATA PROTECTION
The feminist struggles arise as a reaction to the historical subordination of women to men. The objective of feminist theory is to analyse the causes of this subordination and reverse them. It is precisely with the aim of putting an end to women's inferiority of power that feminism brings to light the problems of privacy.
Feminism, maintains Agra Romero (2012), is politics and is political theory. The experience of women has not only influenced and provoked changes at a personal and social level, but also at a theoretical level that pushes for reexamining the domains of the political, and questioning the conventional opposition of the public and the private spheres. Then, feminism should not be reduced to an ideology or a movement.
Following Federici (2004: 13), feminist political theory has confirmed that «to look at history from a feminist viewpoint means to redefine in fundamental ways the accepted historical categories and to make visible hidden structures of domination and exploitation».
«The personal is political» (Hanisch, 1970; expresses the need to overcome the liberal dualism in which the personal becomes depoliticised. It notes that this division leads to an exclusion of the problems that affect the personal from political discussion and action. The claim denounces the inability of privacy to prevent situations of domination in the private sphere, by keeping out of public scrutiny the impunity of those who dominate. It also draws attention to the liberal tendency to look at social life exclusively in personal terms, and alerts about the importance of political actions to overcome domination in the personal sphere. Furthermore, a large part of feminist studies (Fraser, 2012;Hanisch, 1970;Pateman, 1989;Young, 1987) understand that overcoming the depoliticisation of the personal implies pursuing a different social order, based on a social conception of individuality within which public and private dimensions are distinguished, but not opposed. With regards to this, García Manrique (2013), points out the importance of feminist's contributions to the terms in which private life must be sustained by a politically organised community. He notes that, although politics should not have to establish the conditions of the private life, it must become aware that the private sphere depends on publicly imposed conditions. And also that freedom depends, too, on the conditions of the private life. Thus, the political community should not ignore private life, but rather it should enable and promote the conditions so that the private life can also contribute to everyone being equally free.
These analysis and conclusions of the feminist critique of privacy are, from my point of view, applicable to other situations of inequality comparable to those that cause the subordination of women. I mean here to emphasise that the theoretical outcomes of the feminist movement are not only applicable to women's struggles, but also to diverse critical analysis of unequal social structures. As regard to this, the claim «the personal is political» may be, in my view, very illuminating to rethink PDP regulations. It shows us that the solution to the personal vulnerabilities caused by personal data's massive exploitation cannot be restricted to isolated action. Moreover, that, even if the «agreement» between the citizen and a certain company that uses her or his data is a «private» relationship, the protection of personal data should be excluded from public scrutiny.
In that light, equating the right to PDP with the inviolability of personal data blocks the possibilities of tackling the root causes of data domination. By establishing individual consent as the only mechanism citizens have to protect their data, GDPR confines people to individual resistance facing highly concentrated and organised markets, whose use of big data affects the personal affairs of all of us. On doing so, the Regulation cuts off the possibilities of giving a socially organised response to the implications of big data exploitation. In that sense, GDPR replaces social protection by what Harvey (2004) calls a «personal responsibility system». A system devoid of democratic control where individual cessions shape the type of society where we all live. Besides, it confers the market the status of «appropriate guide» for the development of big data technologies. Within this situation, people's options restrict to two: to trust the markets' goodwill; or to act as technophobes, trying to marginalise their own actions and beliefs from data accumulation. This situation entails the risk of entering privacy into a game of preferences that excludes those who do not share the rules imposed by data business. Arruzza, Bhattacharya and Fraser (2019: 50) note that the effect of this system, rooted in the institutional structure of capitalism, «is to declare vast swaths of social life off limits to democratic control and turn them over to direct corporate domination». It fails to fulfil the supposed role of public institutions in contemporary welfare states: to replace the logic of the markets by an equal distribution of goods, through the recognition of a set of legal rights to all members of a political community (Marshall, 1950). On these grounds, it seems clear that we should not seek protection for personal data in darkness, secrecy and from the isolation. Rather, the answer must be, also, political. Moreover, feminism lays bare the need of protecting the personal from the subordination of the individual will, instead than only from its interference.

CHANGING PERSPECTIVE
In my view, the dark side of privacy is a consequence of defining its value in terms of (absence of) interferences. Within this framework, to ensure privacy consists of building walls around the private and the personal to guard it from external interferences, but not from subordination. In the case of PDP, the role of the state restricts to guaranteeing the «self-management» of personal data. That is, that no one could force or prohibit another to consent the unwarranted publication, gathering or use of her or his data. Privacy in that sense lacks aspirations for equality in the private sphere by leaving outside its scope of protection the inequalities of power affecting private relations. This applies not only to the traditional conception of privacy, but also to some of its critics, such as Etzioni (1999), Solove (2003) or Allen (1999), who do not move significantly away from the definition of privacy in terms of noninterference.
Do these weaknesses of the liberal model of privacy mean that privacy is not a suitable value for underpinning PDP regulations? Should they lead us to forget the distinction between the public and the private? Certainly not. As I understand privacy, it is a sphere where we enjoy our freedom in a particularly individualised way. Therefore, the idea of freedom we refer to determines the way we understand privacy and the protection of the private-personal. This idea has been also noted by Roberts (2018: 6): «re-examining the beliefs that we hold about the nature of freedom -its constituent conditions-might, if it causes a shift in those beliefs, lead us to think differently about privacy».
Consequently, my proposal is not to reject or to restrict privacy as the articulating value of PDP regulations, nor to reject or to restrict individual decision as the main mechanism to protect personal data. Rather, it seems necessary to defy the liberal hegemonic conception of privacy by rethinking its value from an alternative conception of freedom.
The aim of this article is not to delve deeper into this issue. Even though, in the following lines I sketch out that a neo-republican conception of privacy would overcome the problems of liberal privacy, and so, it would be may be more suitable for protecting people from data domination.
Neo-republicanism defines freedom as non-domination (Pettit, 1997;. The main difference between liberal and republican freedom lies in the type of hindrances that they consider inimical to freedom. While liberalism considers that there go against freedom all the interferences that modify the individual will that guide a course of action; republicanism understands that a person is not free if he or she is exposed to another's uncontrolled power of interference. Whether if this power affect directly or indirectly the individual capability of choosing a valuable course of action and carrying it out, and whether this power is actually exercised or just potential. By way of explanation, an agent may dominate another without actively interfering in any of her or his actions, as well as it is possible to interfere in the actions of an agent without dominating him or her. In that line, Pettit notes that dominating relationships can have their origin in consent in case it gives one party the uncontrolled capacity to influence the life of the other. Therefore, republicanism recognises the importance of power relations when thinking about freedom. Contrarily to liberal freedom, republican freedom tends to social equity. This implies that preventing one agent from dominating another is a way of ensuring the freedom of all, and not a form of limiting it. To be free in the republican sense means to not being exposed to any power of interference that one cannot autonomously control. In that sense, freedom requires a series of material and immaterial conditions that must be equally accessible by every citizen. The role of public powers is then to protect and resource citizens against domination, provided that this action of the state responds to the mandate of the citizenry (Pettit, 2012). Bertomeu and Domènech (2005) note that «X is strengthened in his civicpolitical freedom by a more or less large core of constitutive (not purely instrumental) rights that no one can take away from him, nor can he himself alienate (sell or donate) at will, without losing his status of free citizen» (our transl.). From this perspective, the liberal dichotomies between freedom and other goods or the public interest, lose their meaning, since relationships between rights become a matter of reciprocal necessity.
Drawing on republican theory, privacy would be closer to a sphere free from arbitrary interferences in order to ensure individual's will concerning private and personal issues, but also her or his ability and means to carry out such will. Privacy would still be a sphere governed by the individual will, shared with the subjects that the person wants and where the individual does what he or she wants to do. But, at the same time, it would allow to guard the individual from isolation when facing situations of oppression at a personal level. Contrarily to liberal privacy, republican privacy would protect the individual not only from violations of her or his individual will regarding private issues, but also from the subordination to others' wills.

CONCLUSIONS
Privacy is a distinctive value of liberalism. It underpins the European legal framework for PDP.
From the lens of liberal theory, privacy is an «inviolable» sphere proof against others' interferences. Ensuring privacy means guarding the individual against the injury inflicted by invasions upon her or his personal affairs. This conception of privacy is unable to protect the individual from the subordination of her or his will, as feminism has shown.
The second-wave feminist claim «the personal is political» pushes for critically examining the traditional opposition between public and private spheres. Feminist theory notes that asymmetric power relations between «private agents» give rise to the oppression of the less powerful agent. By keeping the causes of these inequalities away from public discussion and action, public institutions tend to naturalise the resulting subordination of ones to others and to redefine it as a «personal problem». A clear example is the historical subordination of women to men. On doing so, politics relegates the resolution of these problems to individual fights, complicating to tackle their root causes as a society.
Beyond being a key element in women's struggles, this feminist critique of privacy represents a substantial contribution in theoretical terms. This article shows that it is an illuminating approach to reviewing the meaning and scope of PDP regulations. Looking at PDP from a feminist point of view means: Firstly, to redefine the inviolable conception of privacy that underpins GDPR in order to overcome the depoliticisation of personal data protection. Secondly, to make visible the hidden structures of domination behind the massive exploitation of personal data. Thirdly, to stop thinking that individual action is the only way to freely protect personal data. Society has to use its ability to decide collectively which would be the best way to protect personal data.
Keeping this in mind, my proposal is not to reject or to restrict privacy as the articulating value of PDP regulations. I stand up for defying the liberal conception of privacy in light of the republican conception of freedom as nondomination and so, recognizing the social dimension of the right to PDP. This does not mean one has to force people to keep hidden their personal data, nor to take decisions about the protection of their data in a particular sense. It is about to ensure that individual decisions regarding private issues are not (directly or indirectly) subordinated to others' interests. Consequently, it is critical to reduce the factors that give a few corporations the power to undermine citizens' ability to act autonomously for the protection of their personal data.
This requires that public institutions replace the logic of data business for the logic of fundamental rights, limiting the unbridled expansion of the markets to data. It is essential to promote the widest possible public debate on the development of big data technologies and personal data protection regulations, even questioning its conceptual foundations. Public institution should support this debate by providing mechanisms of transparency, external control and accountability of big data technologies and data-driven corporations. All this would legitimate political actions to ensure everyone's possibilities to detect and protest against unfair big data exploitation practices, and to take free decisions about their personal data away from the pressures of data business.

ACKNOWLEDGMENTS
I would like to thank the organiser and participants of the «Value of Privacy» track at the Amsterdam Privacy Conference 2018 for their comments and questions on an earlier draft of this article.