No Autonomy without Equality!

An interview with Denise Celentano

In Inventing the Future, Nick Srnicek and Alex Williams suggest that a rejuvenated Left will require the installation of a new universal ideal.  Such an ideal acts like a flag at the top of a mountain.  Although it is unreasonable to expect we will ever make it to the summit, it is still useful to look up at the flag, see how far we have to go, and make sure we are climbing in the right direction.

The form that this universal ideal should take is open to debate, and opinions vary from writer to writer (or often remain unexamined, in the background).  Whilst some have promoted some version of autonomy as the ultimate goal, others have argued for social change in the name of concepts like freedom and justice – ideas which would seem to nest and overlap in interesting and complicated ways.  In a more contemporary context, some campaigners on the Left are also arguing for a politics of human flourishing or well-being – ideals which would also seem to have their own advantages and disadvantages as a way of orienting political struggle.

On what ultimate grounds, then, should we argue for social change, and are some grounds perhaps stronger and more valid than others?  This is the key concern taken up by the philosopher Denise Celentano, in her recent paper on ‘Automation, Labour Justice and Equality’.  I spoke to Denise to find out her thoughts on the challenges facing the politics of work today.

David Frayne: Denise, let us begin in the obvious place, with the question of automation.  My impression from your article is that you feel dubious about the narrative we’ve been seeing in the media, around mass automation and the coming ‘job apocalypse’. Can you explain what you find dubious about this narrative?

Denise Celantano: I find this narrative dubious for two main reasons. First, research shows that automation is having an impact on tasks reducible to replicable patterns, rather than entire jobs. One the one hand, this means that not everyone is equally affected by automation, since low-skilled workers are more exposed than others. On the other hand, there is also a logic of tasks being externalised from AI to humans, which has not been sufficiently represented in the debate.  Humans are required to fill in the gaps of AI, to train AI, to integrate AI’s tasks, and to do the jobs that, so far, are too expensive for AI to perform. The picture is more complex than the ‘spectre of joblessness’ can possibly tell. For example, self-driving cars require a – very human – team to train, check, and adjust the process. Lilly Irani talked about ‘data janitors’ to describe the work of taskers that ‘clean the internet’ from inappropriate content, choose the right tag images, and so on (what we might call ‘digital labor’). These are all examples of work hidden behind automation. Of course, what is hidden also lacks of the possibility to become a political concern.


Recall for example the ‘Mechanical Turk’, an 18th century chess-playing machine. Although it was presented as an ‘automaton’, it turned out to be driven by someone – the ‘Turk’ – behind the curtain: it was a fake automaton. Sociologists often use this example to highlight the work performed behind the scenes of technology. There are very interesting studies also on the gendered character of this hidden human labor. Thus, it would be better to talk about the invisible division of labor between AI and humans and of ‘heteromation’ (as suggested by Ekbia and Nardi).  These concepts are more appropriate and useful than mere ‘joblessness’ to address core questions related to the impact of technological change on work.


More broadly, societies are complex and so are their needs. Needs that have historically come to be satisfied by machines do not merely ‘liberate time’ and labor; they also produce new needs. From this perspective, Ivan Illich was forward-looking decades ago: as he observed, although cars were supposed to allow us to save our time, they also produce the traffic jam. Likewise, emails were supposed to make our life easier, and yet for most of us they have become a nightmare! (In France a ‘right to disconnection’ of workers has been recently established to tackle this). To get a grasp on this, we should read the old, and yet impressively timely book of Ruth Schwartz Cowan, More Work for Mother. When washing machines and all sorts of appliances changed the routine of the households, and thus were expected to change the life of women, this did not actually make the division of labor in the household less gendered. Even with plenty of technology in their homes, women are still doing the bulk of the job. Technology cannot be decoupled from the cultural context in which it is embedded.


These examples show that we should not think of the issue of automation as a zero-sum game between us and technology. Focusing all our attention on the “great substitution” (of labour by machines), as David Autor called it, prevents us from seeing urgent and relevant questions. In my research, I am mostly interested in work fairness. Thus I would ask: how can technology help us to rethink the division of labor, so as to ensure that all benefit from automation not only in distributive, but also contributive terms – that is, in their work?


And this leads me to the second reason to question the ‘job apocalypse’ narrative. Even if hyper-sophisticated technologies will change everything, I don’t think that the dispute between joblessness advocates and joblessness sceptics will get us far.  Rather than waiting for joblessness to ‘happen’, as if it was a natural calamity and not directed by humans, we should ask how technology be managed in such a way that all can benefit from it.


So the narrative on the end of work is not only empirically dubious, but it also tends to be politically misleading. It has a certain appeal in public discourse because there is a sort of collective emotional over-investment on the topic, a mix of hopes and fears that do not help the debate to progress. After all, we have been waiting for centuries to be ‘liberated’ from menial labor by technology.  This is a kind of ancestral dream.

DF – If we are not in fact likely to witness the end of work, then what implications does this have for political struggles on the Left? Can you explain, for example, what you mean by the ‘desirability view’ of automation, and why you have questioned it?

DC – I use the ‘desirability view’ to refer to approaches that address automation and work primarily in terms of questions of whether or not it is desirable to work. That is, the desirability view does not ask how work is to be made fairer, or how technology may fit this purpose, rather, it asks ‘should we work?’ Some theorists answer yes, we should, because without work our lives would lack of a chief source of meaning and purpose, we would lack of chances for social esteem, and so on. Other theorists say no, work is just drudgery at the service of a boss and nothing but a way to pay the bills, thus we should get rid of the labourist ideology and de-romanticize work.


I think that both of them are reasonable in some sense, but they are more reasonable when considered jointly. If taken alone, each of them overlooks something important that the other perspective addresses: sometimes theories that emphasize the goods of work tend to convey an idealized picture of work that simply does not match reality, for they overlook inequalities of power, for example. In that case, the post-work ethical narrative may help to ground us back in the real world. On the other hand, the de-romanticized narrative of the refusal of work, while vital in this sense, tends to be paternalistic and one-sided too – and it is at risk of falling into a reversed romanticism, into a sort of ‘ethical bovarism of work’ so to speak. In fact, while dreaming about a jobless future, we still need people to do tasks and we still benefit from the work of others – but we don’t ‘see’ them because we are too immersed in the dream. I mean, the problem of needs is hugely absent in this debate. Likewise, the pleasure of doing things that have meaning for us, the rewarding ‘flow’ we sometimes experience while working, cannot be entirely dismissed as a mere ideological myth. One does not have to be an advocate of the ‘labourist ideology’ to see this. Thus I think that both perspectives are reasonable for different reasons, but that neither of them should be taken as having the last word on the issue. If we consider their insights as complementary, we get to the conclusion that work does have some role in shaping our social identity and as a source of meaning, and yet, work should not monopolize our social identity and sources of meaning.


But let me stress another crucial point here. Although these positions are normally taken as patently opposite, I think that they are closer than it appears. They have in common the very way in which to set the whole problem, as a problem of deciding whether work ‘in itself’ is desirable or not. I am not saying that this is not a relevant question. Whether we have a ‘moral obligation’, as some say, to work is certainly a crucial topic, with key political consequences. For example, the choice between workfare and unconditional income depends on the answer to the question of whether people should work. But these theories alone do not tell us how to make work fairer. They do not tell us who will actually afford leisure and temporal autonomy. They do not tell us who will benefit from automation. What happens once we have crossed the door of work – which still exists, because social needs still exist – is not part of the scope of the desirability question. While we are speculating about whether work is inherently desirable or not, we still want restrooms to be cleaned, our correspondence to be delivered, our laptop to be repaired, our food to be produced and placed on shelves in supermarkets, and so on. Social needs and the work they require cannot be dismissed as a temporary residue of the detestable work culture.


Likewise, theorists who argue for subjective self-realization in the workplace also tend to overlook the fact that some are in a position to outsource the undesired tasks to others, and thus have far more scope for self-realization than others. This ‘contributive advantage’ consists of the possibility to perform more rewarding tasks, thanks to the possibility to outsource the undesired ones to others. In the case of reproductive work, this happens on a global scale, in what Arlie Hochschild called the “global chains of care”: the outsourcing of care work from Western families to migrant women.


Addressing the issue in terms of desirability, rather than in terms of justice, does not allow us to address these situations. So I think that we should set a sort of moral priority of questions of justice in work structures, over and above questions of desirability. That is, before asking whether work ‘in itself’ is desirable or not, let’s make sure that the existing division of labor be fairer. Otherwise there is the serious risk that the desirability question becomes an exclusivist question, which means, a question that only few can actually afford address.

DF – One of the things I really value about your work is its sober discussion of the Universal Basic Income. Whilst the recent enthusiasm for the idea has been wonderful, I think there has been a tendency to perceive the policy as something of a silver bullet.  In your article, you question whether UBI is really enough to secure labour justice.  Why do you doubt the idea that labour justice ‘flows’ from UBI?  What else does labour justice require?

DC – The debate around UBI is a good thing for a number of reasons. It has helped in shaping and legitimizing an alternative normative language to address the irrationality of a system with chronic underemployment and exploitation, which keeps patronizing people, telling them that they are morally obliged to work and blaming them if they don’t, while using work as the exclusive means to allocate resources. The issue of economic blackmail is crucial, and it is striking to see how many theorists argue for free occupational choice while rejecting income unconditionality. UBI has helped to give visibility and resonance to this crucial issue.


However, while acknowledging the important role of UBI in the debate, it seems to me that there is a tendency to delegate the whole burden of justice to UBI. We are asking UBI to fix too many things. For example, with respect to work, UBI advocates often say that UBI allows for an ‘exit choice’ – the ability to leave an undesirable job –  and would also provide workers with more bargaining power.  This way, UBI is said to benefit labor justice.  In my article I have called this a ‘ripple effect’ conception of justice: by fixing income, work fairness is expected to follow in an almost automatic way. But UBI is an allocative response to a far more complex question. To be sure, questions like who will take the lion’s share of technological progress in terms of profits are of the highest importance. But it is equally important to acknowledge ‘contributive’ concerns. From the perspective of UBI, for example, a phenomenon like occupational segregation (which technology does not solve) is not really made visible. Likewise, we should not forget that in principle UBI is compatible with the most authoritarian forms of government. The ‘exit choice’ is not the same as democracy. And we should not forget that UBI is essentially a First World option. What about the countries where we keep outsourcing labor at lower costs?


I am not against UBI, but if I had to decide, I would favour a critically contextualized form of unconditionality. Not a UBI that – like an allocative deus ex machina – fixes all our problems, and to which to delegate the whole burden of justice, but a critically cautious unconditionality – one that is not intended to replace things like economic democracy, welfare provisions, and labor justice.

DF – Is it possible to summarise, briefly, the normative ideal you think would best orient a politics of work?

DC – I think that if work and the division of labor are complex things, so should be our normative ideals. UBI is a response that addresses only one dimension: income. But there are many dimensions involved in work. There is a relational dimension, a political dimension, and what I have been calling a contributive dimension, which regards the nature of work tasks and occupations, along with labor time. For decades theorists of justice have restricted their attention to the economic side of the issue, overlooking a number of other concerns, like the way people treat each other, or the very way we organize work.  Status, for example, has sometimes remained outside of the picture, even though it is deeply involved in the dynamics of exploitation.


For this reason, in my article I have suggested a ‘multidimensional’ ideal of labor justice, which I have called ‘contributive parity’. I came to this concept because I realized that in the literature, very few authors address work in terms of equality, and that concepts of autonomy and freedom are not sufficient in themselves, because they are compatible with occupational segregation and contributive subordination.


One advantage of the ideal of contributive parity is that it is not paternalistic. Unlike some approaches, contributive parity does not tell people what they should think about work. It does not tell people what they should feel about work, it does not claim to have identified the very ‘essence of work’ and its ultimate role in people’s lives. I do not want philosophers to tell me whether work is the most important thing in my life, or whether I should better devote my energies to leisure. This is all my business. I want to be able to negotiate my own way of experiencing life, work, and the value to attribute to them, without philosophers telling me that if I enjoy working I’m a victim of the workist ideology, or telling me that if I devote my energies to leisure there’s something wrong with me. From this perspective, contributive parity has a crucial liberal strand. It says that regardless of the way we interpret the ultimate meaning of work, work should be fair, and that the costs of the self-realization of some should not be paid by others (usually people from the most socially vulnerable groups).


This ideal says that no one should be prevented from contributing to social cooperation as a peer, and that in order for this to be possible, four key dimensions of justice should be satisfied: economic, because without freedom from material need one cannot be free to choose; relational, because no one should be treated as an inferior in the division of labor; political, because no one should be powerless at work; and contributive, because no one should be reserved for the less interesting, routine tasks, with low self-direction, low prospects, and heteronomy.


Contributive parity is multidimensional because it says that in order for us to be genuine peers in social cooperation, these conditions are to be satisfied jointly. That is, one cannot have contributive parity only by fixing the income side of the issue (even if this is crucial). Nor can one have contributive parity only by claiming that all should be treated as equals at work, without this leading to more structural concerns, whereby status inferiority is reproduced in the very way that work is organized. Likewise, one can hardly have contributive parity realized if she does not have a voice in the decision-making processes that affect her work. And certainly, contributive parity is not realized when certain tasks and occupations are reserved for the least advantaged – under a regime that legally rejects occupational segregation, but that still has a disproportionate number of migrants and women performing the tasks at the lowest rung of the occupational ladder.


I know that it may sound utopian, and it certainly is, even though I do not use the term ‘utopian’ in pejorative terms. Technically speaking, utopias are places that do not exist yet, and nonetheless in one sense they do already exist in people’s discourses.  Publicly circulating discourses spread meanings and interpretations that produce effects on the real world, on the choices we make, on what sounds like a ’legitimate claim’ and what does not – they have thus a ‘performative’ side in some way. Thus, let me point to this ideal as a potentially performative one. Yet let me also suggest that this ideal is not the mere product of intellectual speculation, but can be extrapolated from existing social struggles around work and its organization. Struggles for a fairer division of labor underlie normative demands for equality, not only for autonomy.

DF – What is your opinion of some of the other ideals that have been used to orient the politics of work?  Isn’t it perhaps the case that ideals like ‘autonomy’ and ‘freedom’ already have ideas about equality and justice built into them?

DC – In the history of political thought and of social struggles one finds plenty of ideals: self-realization, flourishing, freedom, and so on. The ideal of autonomy is, it seems to me, the most widespread in current debates. When one reviews the literature about work, one finds that the only norms at our disposal, in most cases, are that of freedom and autonomy, even if these are interpreted with lots of different inflections and meanings. It is striking to see that even completely different views converge on the importance of autonomy. Both theories of meaningful work and post-work ethics, for example, despite their deep differences, ground their normative views on work in ideals of autonomy and freedom.


Of course, I am a fan of freedom and autonomy. But I am concerned about the fact that without equality, autonomy is at risk of becoming an exclusivist ideal. Plenty of people experience free occupational choice and autonomy in the workplace, and yet, they also happen to be the most advantaged people. They can be free because someone else will do the job that they are allowed not to do. With the only ideal of freedom to hand, we might overlook that the scope of some people to refuse work is often grounded in the unfree work of others.


A liberation from work ‘for the few’ is clearly not the ideal we should defend. Meaningful work for the few, or the refusal of work for the few: is that really what we want? Taken alone, I think autonomy is an incomplete ideal.  So no, I don’t believe that equality and justice are already somewhat ‘internal’ to the ideal of autonomy. We cannot take them for granted or assume they are implicit. They should be named and defended in their own right.

DF – In the UK there is now a newer trend of orienting struggles according to the politics of well-being. The idea is that we should focus on promoting those policies that offer the most robust response to problems like work-related stress, or the psychological misery caused by poverty and social exclusion.  Although I know your work does not reflect on the concept of well-being, do you perceive any potential strengths or weaknesses in building a politics of work on this kind of platform?

DC – There is an Italian movie, Paolo Virzi’s Tutta la Vita Davanti (Your Whole Life Ahead of You), where workers in a call center, living extremely precarious lives under the blackmail of economic insecurity, are taught a ‘motivational dance’ that are required to do at the beginning of every working day. The underlying assumption is that if they feel alienated, anxious, oppressed at work, this is because there is something wrong in their attitude. They do not feel the ‘right feelings’. This psychologization of labor injustice goes along with an individualized narrative that attributes the failures of the system to the personal traits of individuals. According to this view, individuals are lacking in some mysterious ability to tolerate a system that demands they be happy, while at the same time making happiness impossible.


I am more familiar with the Italian and French context, but I think that the category of ‘wellbeing’ should not monopolize what we can possibly say and do about making work a more human, fairer thing. I would say that addressing the problem of labor justice in terms of ‘wellbeing’ hardly bothers anyone. It does not challenge existing hierarchies and power inequalities. One can feel good after the ‘motivational dance’ in the morning, and yet still occupy the lowest rung of the ladder, or fae the constant threat of losing the job. There is a philosophical figure that well describes this paradox: the ‘happy slave’. Addressing labor injustice on the ground of individual psychologies, without contextualizing them into the systemic reproduction of inequalities, allows us to blame individuals for the failures of the system, while absolving institutions from all responsibility.

DF – When I stepped back from your article, I could definitely see more clearly the drawbacks of making debates on the future of work too much about ethics (questions of how much one should value work, what sort of values can replace the work ethic, and so on). However, although I definitely agree that questions of equality and justice need to be foregrounded, I would be reluctant to turn away from ethics.  You might partly attribute this to my being from the UK, where the government’s aggressively pro-work ideology is having a disastrous impact on groups like disabled people.  My question then: are you saying that we should turn away completely from ethical questions about desirability, and the value of work?

DC – No, I am not. Even though my very initial stance was to refuse these approaches, at a closer look I realized that we definitely need a conversation on the ethics of work. This conversation is vital precisely for the reason you mentioned. The way we publicly tell the story of work, so to speak, deeply affects choices and policies and ultimately people’s lives. Public discourses convey interpretations about the legitimacy and the desirability of claims and worldviews. Discourses where everyone believes that without work your life is worthless, for example, powerfully impact what a government can possibly do, and what people are ready to accept as a sound intervention.


If we grow up in a world where the ‘work ethic’ prevails in public discourses and common sense, we are far less likely to see the possibility for alternatives, and far more likely to apportion blame on those who embrace different views. This impoverishes the normative languages at our disposal to imagine and struggle for a different world. So yes, ethics matter.  They matter for how we tell the story of work in our publicly circulating discourses, and this has real consequences for people’s lives. Moreover, having a conversation around the ethical value of work and leisure helps cast light on the underexplored zones of our lives, and nurtures our social self-understanding as individuals living in complex, changing societies. Since ethics is inherently plural, by definition there are plural ways in which we can interpret the value of work. I see this ethical pluralism about work as totally healthy and necessary. Without different ethics there would be no democracy, no social dynamism, no social change, and no freedom. Of course, it is a matter of political struggle as to which ethics of work prevails. The boundaries between competing worldviews on work are continuously subject to public scrutiny and contestation.


Another fundamental contribution of ethical discourses about work is that they provide us with insights about relationships, feelings, emotions, and values. At the end of the day, these are the things that most of us really care about. For example, I recently came across an old interview with Roland Barthes in Le Monde, focussing on idleness as a tactic against structures of constraints. I know that there is now an attempt to rehabilitate the ethics of idleness and I really enjoyed reading that interview.  And yet, when it comes to the politics of work, we cannot restrict our attention to the cultural nuances of the value of idleness. We should also ask: is the kind of free-time being promoted really for everybody?


In the past, idleness was a matter of class. As Veblen famously showed in his book on the ‘leisure class’, it was a mark of social distinction.  But today things are different. The relationship is reversed: the ‘leisure class’ is now the ‘busy class’. Recent research has shown that being busy is the new mark of distinction, replacing leisure in the complex geography of status. Cool people don’t have time.  Idleness is not cool. From this perspective, the post-work ethical narrative can be understood as a healthy counter-discourse to the current hegemony of being busy and working hard. But without justice, that of the refusal of work is at risk of being an ethics for the few.  Overall, I maintain that any ethics of work/non-work cannot be decoupled from concerns for justice. And that the latter should have a sort of moral priority over the former.

DF – Are there any emerging struggles or trends that make you hopeful for the future?

We have recently seen the struggles of Uber workers, cleaners, and broader transnational protests of the feminist movement, focussing on the problem of reproductive labour. Women, young people, people in poverty, migrants, and racialized people are those in the best position to see the failures of the system and of the current crisis of work. I like to imagine a sort of global struggle for work fairness where people who experience contributive subordination keep up the pressure to change the rules of the game. Guy Standing has identified the precariat as a possible leading force for such a struggle, but the decline of Left movements throughout the world is concerning. Regarding automation, there are valuable initiatives coming out of academia. Consider for example the ‘Fair Work’ initiative led by Mark Graham, which uses the reputational power of certification as a tool to put pressure on companies to stop exploiting digital workers.


But I am also concerned with the struggles that don’t emerge in disruptive protests because they don’t have distinct unions, or a political, normative code to mobilize. I am concerned by suffering that remains confined to the individual’s private solitude, and that leads to depression and sometimes suicide. This is another reason to bring public attention to these themes. There is a social suffering around work that is still striving to be represented, still striving to turn into a real struggle.

DF – Can you recommend Autonomy a book or a film you have recently enjoyed on the politics of work?

DC – A book that I would definitely recommend is Private Government, by Elizabeth Anderson. It’s historically informed, theoretically sharp, politically engaged, and even funny to read. A film? Maybe the most appropriate recommendation here would be Hidden Figures. It tells the story of black women doing the math that led NASA astronauts to the moon. This is a striking example of how technological progress and social backwardness can co-exist, and of how technological progress without social justice is a poor ideal.

Denise Celentano is a Berggruen Fellow at the Department of Philosophy, New York University.

David Frayne is a research affiliate of Autonomy, the author of The Refusal of Work (2015) and is a Berggruen Fellow at the Department of Philosophy, New York University. He can be found on Twitter: @TheWorkDogma