Big Thinker: Michel Foucault

Big Thinker: Michel Foucault
ArticleBig Thinkers + Explainers
BY The Ethics Centre 13 JUL 2017
Michel Foucault (1926—1984) was a French philosopher, historian and psychologist whose work explored the underlying power relationships in a range of our modern institutions.
Given Foucault’s focus on the ways institutions wield power over us, and that trust in institutions is catastrophically low around the world today, it’s worth having a look at some of the radical Frenchman’s key ideas.
History has no rhyme or reason
At the centre of Foucault’s ideas is the concept of genealogy – the word people usually use when they’re tracing their family history. Foucault thought all of history emerged in the same way a family does – with no sense of reason or purpose.
Just like your existence was the result of a bunch of random people meeting and procreating over generations, he thought our big ideas and social movements were the product of luck and circumstance. He argued what we do is both a product of the popular ways of thinking at the time (which he called rationalities) and the ways in which people talked about those ideas (which he called discourses).
Today Foucault might suggest the dominant rationalities were those of capitalism and technology. And the discourse we use to talk about them might be economics because we think about and debate things in terms their usefulness, efficiency and labour saving. Our judgements about what’s best are filtered through these concepts, which didn’t emerge because of any conscious historical design, but as random accidents.
You might not agree with Foucault. There are people who believe in moral progress and the notion our world is improving as time goes on. However, Foucault’s work still highlights the powerful sense in which certain ideas can become the flavour of the month and dominate the way we interpret the world around us.
For example, if capitalism is a dominant rationality, encouraging us to think of people as economic units of production rather than people in their own right, how might that impact the things we talk about? If Foucault is right, our conversations would probably centre on how to make life more efficient and how to manage the demands of labour with the other aspects of our life. When you consider the amount of time people spend looking for ‘life hacks’ and the ongoing discussion around work/life balance, it seems like he might have been on to something.
But Foucault goes further. It’s not just the things we talk about or the ways we talk about them. It’s the solutions we come up with. They will always reflect the dominant rationality of the time. Unless we’ve done the radical work of dismantling the old systems and changing our thinking, we’ll just get the same results in a different form.
Care is a kind of control
Although many of Foucault’s arguments were new to the philosophical world when he wrote them, they were also reactionary. His work on power was largely a response to the tendency for political philosophers to see power only as the relationship between the sovereign and the citizen – or state and individual. When you read the works of social contract theorists like Thomas Hobbes and Jean Jacques Rousseau, you get the sense politics consists only of people and the government.
Foucault challenged all this. He acknowledged lots of power can be traced back to the sovereign, but not all of it can.
For example, the rise of care experts in different fields like medicine, psychology and criminology creates a different source of power. Here, power doesn’t rest in an ability to control people through violence. It’s in their ability to take a person and examine them. In doing so, the person is objectified and turned into a case (we still read cases in psychiatric and medical journals now). This puts the patients under the power of experts who are masters of the popular medical and social discourses at the time.
What’s more, the expert collects information about the patient’s case. A psychiatrist might know what motivates a person’s behaviour, what their darkest sexual desires are, which medications they are taking and who they spend their personal time with. All of this information is collected in the interests of care but can easily become a tool for control.
A good example of the way non-state groups can be caring in a way that creates great power is in the debate around same sex marriage. Decades ago, LGBTI people were treated as cases because their sexual desires were medicalised and criminalised. This is less common now but experts still debate whether children suffer from being raised by same sex parents. Here, Foucault would likely see power being exercised under the guise of care – political liberties, sexuality and choice in marital spouse are controlled and limited as a way of giving children the best opportunities.
Prison power in inmate self-regulation
Foucault thought prisons were a really good example of the role of care in exercising power and how discourses can shape people’s thinking. They also reveal some other unique things about the nature of power in general and prisons more specifically.
In Discipline and Punish: The Birth of the Prison, Foucault observed a monumental shift in how society dealt with crime. Over a few decades, punishments went from being public, violent spectacles like beheadings, hangings and mutilations to private, clinical and sterile exercises with the prison at the centre of it all. For Foucault, the move represented a shift in discourse. Capital punishment and torture were out, discipline and self-regulation were in.
He saw the new prison, where the inmates are tightly managed and regulated by timetables – meal time, leisure time, work time, lights out time – as being a different form of control. People weren’t in fear of being butchered in the town square anymore. The prison aimed to control behaviour through constant observation. Prisoners who were always watched, regulated their own behaviour.
This model of the prison is best reflected in Jeremy Bentham’s concept of the panopticon. The panopticon was a prison where every cell is visible from a central tower occupied by an unseen guard. The cells are divided by walls so the prisoners can’t engage with each other but they are totally visible from the tower at all times. Bentham thought – and Foucault agreed – that even though the prisoners wouldn’t know if they were being watched at any moment, knowing they could be seen would be enough to control their behaviour. Prisoners were always visible while guards were always unseen.
Foucault believed the panopticon could be recreated as a factory, school, hospital or society. Knowing we’re being watched motivates us to conform our behaviours to what is expected. We don’t want to be caught, judged or punished. The more frequently we are observed, the more likely we are to regulate ourselves. The system intensifies as time goes on.
Exactly what ‘normality’ means will vary depending on the dominant discourse of the time but it will always endeavour to reform prisoners so they are useful to society and to the powerful. That’s why, Foucault argued, prisoners are often forced to do labour. It’s a way of taking something society sees as useless and making it useful.
Even when they’re motivated by care – for example, by the belief that work is good for prisoners and helps them reform – the prison system serves the interests of the powerful in Foucault’s eyes. It will always reflect their needs and play a role in enforcing their vision of how society should be.
You needn’t accept all of Foucault’s views on prisons to see a few useful points in his argument. First, prisons haven’t been around forever. There are other ways of dealing with crime we could use but choose not to. Why do we think prisons are the best? What are the beliefs driving that judgement?
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
James C. Hathaway on the refugee convention
James C. Hathaway on the refugee convention
VideoLaw + Human Rights
BY James C. Hathaway 22 JUN 2017
To mark the 2017 anniversary of the UN’s World Refugee Day and Australia’s Refugee Week, we spoke to Professor James Hathaway, the world’s leading expert on refugee law.
Watch this video for his in-depth analysis of the Refugee Convention, an agreement made in 1951 in response to World War II. Is it still relevant? What needs to change? He doesn’t hold back!
James’ work is used by courts all over the world when interpreting the Refugee Convention and applying it to decisions. His many publications include the seminal book co-authored with Australia’s Michelle Foster, The Law of Refugee Status.
Follow James Hathaway on Twitter here: @JC_Hathaway
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY James C. Hathaway
James Hathaway is an American-Canadian scholar of international refugee law and related aspects of human rights and public international law. His work is regularly cited by the most senior courts of the common law world, and has played a pivotal role in the evolution of refugee studies scholarship.
Join the conversation
Is there merit in the refugee convention?
Ethics Explainer: Rights and Responsibilities

Ethics Explainer: Rights and Responsibilities
ArticleBig Thinkers + Explainers
BY The Ethics Centre 2 JUN 2017
When you have a right either to do or not do something, it means you are entitled to do it or not.
Rights are always about relationships. If you were the only person in existence, rights wouldn’t be relevant at all. This is why rights always correspond to responsibilities. My rights will limit the ways you can and can’t behave towards me.
Legal philosopher Wesley Hohfeld distinguished between two sets of rights and responsibilities. First, there are claims and duties. Your right to life is attached to everyone else’s duty not to kill you. You can’t have one without the other.
Second, there are liberties and no-claims. If I’m at liberty to raise my children as I see fit it’s because there’s no duty stopping me – nobody can make a claim to influence my actions here. If we have no claim over other people’s liberties, our only duty is not to interfere with their behaviour.
But your liberty disappears as soon as someone has a claim against you. For example, you’re at liberty to move freely until someone else has a claim to private property. Then you have a duty not to trespass on their land.
It’s useful to add into the mix the distinction between positive and negative rights. If you have a positive right, it creates a duty for someone to give you something – like an education. If you have a negative right, it means others have a duty not to treat you in some way – like assaulting you.
All this might seem like tedious academic stuff but it has real world consequences. If there’s a positive right to free speech, people need to be given opportunities to speak out. For example, they might need access to a radio program so they can be heard.
By contrast, if it’s a negative claim right, nobody can censor anyone else’s speech. And if free speech is a liberty, your right to use it is subject to the claims of other. So if other people claim the right not to be offended, for example, you may not be able to speak up.
There are a few reasons why rights are a useful concept in ethics.
First, they are easy to enforce through legal systems. Once we know what rights and duties people have, we can enshrine them in law.
Second, rights and duties protect what we see as most important when we can’t trust everyone will act well all the time. In our imperfect world, rights provide a strong language to influence people’s behaviour.
Finally, rights capture the central ethical concepts of dignity and respect for persons. As the philosopher Joel Feinberg writes:
Having rights enables us to “stand up like men,” to look others in the eye, and to feel in some fundamental way the equal of anyone. To think of oneself as the holder of rights is not to be unduly but properly proud, to have that minimal self-respect that is necessary to be worthy of the love and esteem of others.
Indeed, respect for persons […] may simply be respect for their rights, so that there cannot be the one without the other; and what is called “human dignity” may simply by the recognizable capacity to assert claims.
Feinberg suggests rights are a manifestation of who we are as human beings. They reflect our dignity, autonomy and our equal ethical value. There are other ways to give voice to these things, but in highly individualistic cultures, what philosophers call “rights talk” resonates for two reasons: individual freedom and equality.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
What is more important: rights or responsibilities?
Big Thinker: Hannah Arendt

Big Thinker: Hannah Arendt
ArticleBig Thinkers + Explainers
BY The Ethics Centre 28 FEB 2017
Johannah “Hannah” Arendt (1906 – 1975) was a German Jewish political philosopher who left life under the Nazi regime for nearby European countries before settling in the United States. Informed by the two world wars she lived through, her reflections on totalitarianism, evil, and labour have been influential for decades.
We are still learning from this seminal political theorist. Her book The Origins of Totalitarianism sold out on Amazon in 2017, more than 60 years after it was first published.
Evil doesn’t need malicious intentions
Arendt’s most well known idea is “the banality of evil”. She explored this in 1963 in a piece for The New Yorker that covered the trial of a Nazi bureaucrat who shared the first name of Hitler, Adolf Eichmann. This later became a book called Eichmann in Jerusalem: Reflections on the Banality of Evil.
In Eichmann, Arendt found a man whose greatest crime was a lack of thinking. His role was to transport Jewish people from German occupied areas to concentration camps in Poland. Eichmann did not kill anyone first hand. He was not involved in designing Hitler’s final solution. But he oversaw the trains that took millions of people to their deaths. They were gassed in chambers or died along the way or in camps due to starvation, overwork, illness, cold, heat, or brutality. Eichmann’s only defence for his involvement in this atrocity was obedience to the law and fulfilling his duty.
Eichmann was so steadfast with this line of reasoning he even referenced the philosopher Immanuel Kant and his theory of moral duty. Kant argued morality was acting on your obligations, not your emotions or what will bring you benefit. For Kant, the person who helps the beggar out of empathy or a belief assisting is a pathway to heaven is not doing an act of good. Kant felt everyone is morally obliged to help the beggar, and they are especially virtuous if they act on this duty despite feeling repulsed or no rewarding sense of doing good.
This does not really suit Eichmann’s argument because Kant was emphasising our ability to reason above emotion. This is precisely where Eichmann failed. We can only guess but it seems likely he did his job without asking questions while feeling a sense of comfort in the safety of his salary and senior position during volatile times.
Arendt believed it was this lack of true thinking and questioning that paved the way to genocide. Evil on the scale of Nazism required far more Adolf Eichmanns than Adolf Hitlers.
Totalitarianism needs political apathy
In studying the causes of WWII, Arendt came across “the masses”. She believed totalitarian regimes needed this to succeed.
By “the masses”, she meant the enormous group of people who are politically disconnected from other members of society. They don’t identify with a particular class, religion, or group. Their lack of group membership deprives them of common interests to demand from government. These people have no interest in politics because they don’t have political clout. They are an unorganised cohort with different, often conflicting desires whose needs are easily disregarded by politicians.
But although these people take no active interest in politics they still hold expectations for the state. If politicians fail those expectations they face “the loss of the silent consent and support of the unorganised masses”. In response, they “shed their apathy” and look for an outlet “to voice their new violent opposition”.
The totalitarian leader emerges from this “structureless mass of furious individuals”. With the political apathy of the masses turned to hostility, leaders will rise by breaking established norms and ignoring the way politics is usually done. Arendt dramatically says they prefer “methods which end in death rather than persuasion”. In short, they’re less likely to build politics up than they are to tear it down because that’s what the masses want.
If this all sounds depressing, there is a solution embedded in Arendt’s writing: political engagement. The masses arise when individuals are lonely and politically disconnected. They are defined by a lack of solidarity or responsibility with other citizens.
By revitalising our political community we can recreate what Arendt sees as good politics. This is when people feel a sense of personal and political responsibility for the nation and are able to band together with other citizens who have common interests. When citizens are connected in solidarity with one another, the mass never occurs and totalitarianism is held at bay.
When work defines you, unemployment is a curse
Not all Arendt’s work was concerned with war and totalitarianism. In The Human Condition, she also offers a general critique of modernity. Drawing on Karl Marx, Arendt thought the industrial age transformed humanity from thinkers into “working animals”.
She thought most people had come to define themselves by their work – reducing themselves to economic robots. Although it’s not a central point of Arendt’s analysis, this reduction is a product of the same forces she sees in the banality of evil. It’s a triumph of ‘doing’ over ‘thinking’ and of humans finding easy ways to define themselves.
Arendt wasn’t just concerned because people were reducing themselves to working drones. She also worried the industrial age which had just redefined them was also about to rob them of their new identities. She believed within a few decades, technology would replace factory jobs. Many people’s work would vanish.
“What we are confronted with is the prospect of a society of labourers without labour, that is, without the only activity left to them. Surely, nothing could be worse.”
In a time when automation now threatens over half of jobs on average in OECD countries, Arendt’s predictions seem timely. Can we shift our identity away from work in time to survive the massive job reduction to come?
Published February 2017. Updated August 2018.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
What leads people to support genocide?
Ethics Explainer: Dignity

Ethics Explainer: Dignity
ArticleBig Thinkers + Explainers
BY The Ethics Centre 19 JAN 2017
When we say someone or something has dignity, we mean they have worth beyond their usefulness and abilities. To possess dignity is to have absolute, intrinsic and unconditional value.
The concept of dignity became prominent in the work of Immanuel Kant. He argued objects can be valuable in two different ways. They can have a price or dignity. If something has a price, it is valuable only because it is useful to us. By contrast, things with dignity are valued for their own sake. They can’t be used as tools for our own goals. Instead, we are required to show them respect. For Kant, dignity was what made something a person.
Dignity through the ages
Beliefs about where dignity comes from vary between different philosophical and religious systems. Christians believe humans have dignity because they’re made in the image of God. This is called imago dei. Kant believed humans possessed dignity because they’re rational. Others believe dignity is a way of recognising our common humanity. Some say it’s a social construct we created because it’s useful. Whatever its origin, the concept has become influential in political and ethical discourse today.
A question of human rights
Dignity is often seen as a central notion for human rights. The preamble to the Universal Declaration of Human Rights recognises the “inherent dignity” of “all members of the human family”. By recognising dignity, the Declaration acknowledges ethical limits to the ways we can treat other people.
Kant captured these ethical limits in his idea of respect for persons. In every interaction with another person we are required to treat them as ends in themselves rather than tools to achieve our own goals. We fail to respect people when we treat them as tools for our own convenience or don’t give adequate attention to their needs and wishes.
When it comes to practical matters, it’s not always clear what ‘dignity and respect for persons’ require us to do. For example, in debates around assisted dying (also called assisted suicide or euthanasia) both sides use dignity to argue for opposing conclusions.
Advocates believe the best way to respect dignity is by sparing people from unnecessary or unbearable suffering, while opponents believe dignity requires us never to intentionally kill someone. They claim dignity means a person’s value isn’t diminished by pain or suffering and we are ethically required to remind the patient of this, even if the patient disagrees.
Who makes the rules?
There are also disputes about exactly who is worthy of dignity. Should it be exclusive to humans or extended to animals? And do all animals possess intrinsic value and dignity or just specific species? If animals do have dignity, we’re required to treat them with same respect we afford our fellow human beings.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
How do you measure your worth?
Ethics Explainer: The Harm Principle

Ethics Explainer: The Harm Principle
ArticleBig Thinkers + Explainers
BY The Ethics Centre 27 OCT 2016
The harm principle says people should be free to act however they wish unless their actions cause harm to somebody else.
The principle is a central tenet of the political philosophy known as liberalism and was first proposed by English philosopher John Stuart Mill.
The harm principle is not designed to guide the actions of individuals but to restrict the scope of criminal law and government restrictions of personal liberty.
For Mill – and the many politicians, philosophers and legal theorists who have agreed with him – social disapproval or dislike (“mere offence”) for a person’s actions isn’t enough to justify intervention by government unless they actually harm or pose a significant threat to someone.
The phrase “Your freedom to swing your fist ends where my nose begins” captures the general sentiment of the principle, which is why it’s usually linked to the idea of “negative rights”. These are demands someone not do something to you. For example, we have a negative right to not be assaulted.
On the other hand, “positive rights” demand that others do certain things for us, like provide healthcare or treat us with basic respect. For this reason, the principle is often used in political debates to discuss the limitations of state power.
There’s no issue with actions that are harmful to the individual themselves. If you want to smoke, drink, or use drugs to excess, you should be free to do so. But if you get behind the wheel of a car while under the influence, pass second-hand smoke onto other people, or become violent on certain drugs, then there’s good reason for the government to get involved.
Attempting to define harm
The sticking point comes in trying to define what counts as harmful. Although it might seem obvious, it’s actually not that easy. For example, if you benefit by winning a promotion at work while other applicants lose out, does this count as being harmful to them?
Mill would argue no. He defines harms as wrongful setbacks to interests to which people have rights. He would argue you wouldn’t be harming anyone by winning a promotion because although their interests are set back, no particular person has a right to a promotion. If it’s earned on merit, then it’s fair. “May the best person win”, so to say.
A more difficult category concerns harmful speech. For Mill, you do not have the right to incite violence – this is obviously harmful as it physically hurts and injures. However, he says you do have the right to offend other people – having your feelings hurt doesn’t count as harm.
Recent debates have questioned this and claim that certain kinds of speech can be as damaging psychologically as a physical attack – either because they’re personally insulting or because they entrench established power dynamics and oppress minorities.
Importantly, Mill believed the harm principle only applied to people who are able to exercise their freedom responsibly. For instance, paternalism over children was acceptable since children are not fully capable of responsibly exercising freedom, but paternalism over fully autonomous adults was not.
Unfortunately, he also thought these measures were appropriate to use against “barbarians”, by which he meant non-Europeans in British colonies like India.
This highlights an important point about the harm principle: the basis for determining who is worthy or capable of exercising their freedom can be subject to personal, cultural or political bias. When making decisions about rights and responsibilities, we should be ever careful about the potential biases that inform who or what we apply them to.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
How do you determine what is harmful to another?
Ethics Explainer: Social Contract

Ethics Explainer: Social Contract
ArticleBig Thinkers + Explainers
BY The Ethics Centre 31 AUG 2016
Social contract theories see the relationship of power between state and citizen as a consensual exchange. It is legitimate only if given freely to the state by its citizens and explains why the state has duties to its citizens and vice versa.
Although the idea of a social contract goes as far back as Epicurus and Socrates, it gained popularity during The Enlightenment thanks to Thomas Hobbes, John Locke and Jean-Jacques Rousseau. Today the most popular example of social contract theory comes from John Rawls.
The social contract begins with the idea of a state of nature – the way human beings would exist in the world if they weren’t part of a society. Philosopher Thomas Hobbes believed that because people are fundamentally selfish, life in the state of nature would be “nasty, brutish and short”. The powerful would impose their will on the weak and nobody could feel certain their natural rights to life and freedom would be respected.
Hobbes believed no person in the state of nature was so strong they could be free from fear of another person and no person was so weak they could not present a threat. Because of this, he suggested it would make sense for everyone to submit to a common set of rules and surrender some of their rights to create an all-powerful state that could guarantee and protect every person’s right. Hobbes called it the ‘Leviathan’.
It’s called a contract because it involves an exchange of services. Citizens surrender some of their personal power and liberty. In return the state provides security and the guarantee that civil liberty will be protected.
Crucially, social contract theorists insist the entire legitimacy of a government is based in the reciprocal social contract. They are legitimate because they are the only ones the people willingly hand power to. Locke called this popular sovereignty.
Unlike Hobbes, Locke thought the focus on consent and individual rights meant if a group of people didn’t agree with significant decisions of a ruling government then they should be allowed to join together to form a different social contract and create a different government.
Not every social contract theorist agrees on this point. Philosophers have different ideas on whether the social contract is real, or if it’s a fictional way to describe the relationship between citizens and their government.
If the social contract is a real contract – just like your employment contract – people could be free not to accept the terms. If a person didn’t agree they should give some of their income to the state they should be able to decline to pay tax and as a result, opt out of state-funded hospitals, education, and all the rest.
Like other contracts, withdrawing comes with penalties – so citizens who decide to stop paying taxes may still be subject to punishment.
Other problems arise when the social contract is looked at through a feminist perspective. Historically, social contract theories, like the ones proposed by Hobbes and Locke, say that (legitimate) state authority comes from the consent of free and equal citizens.
Philosophers like Carole Pateman challenge this idea by noting that it fails to deal with the foundation of male domination that these theories rest on.
For Pateman the myth of autonomous, free and equal individual citizens is just that: a myth. It obscures the reality of the systemic subordination of women and others.
In Pateman’s words the social contract is first and foremost a ‘sexual contract’ that keeps women in a subordinate role. The structural subordination of women that props up the classic social contract theory is inherently unjust.
The inherent injustice of social contract theory is further highlighted by those critics that believe individual citizens are forced to opt in to the social contract. Instead of being given a choice, they are just lumped together in a political system which they, as individuals, have little chance to control.
Of course, the idea of individuals choosing not to opt in or out is pretty impractical – imagine trying to stop someone from using roads or footpaths because they didn’t pay tax.
To address the inherent inequity in some forms of social contract theory, John Rawls proposes a hypothetical social contract based on fundamental principles of justice. The principles are designed to provide a clear rationale to guide people in choosing to willingly agree to surrender some individual freedoms in exchange for having some rights protected. Rawls’ answer to this question is a thought experiment he calls the veil of ignorance.
By imagining we are behind a veil of ignorance with no knowledge of our own personal circumstances, we can better judge what is fair for all. If we do so with a principle in place that would strive for liberty for all at no one else’s expense, along with a principle of difference (the difference principle) that guarantees equal opportunity for all, as a society we would have a more just foundation for individuals to agree to a contract that in which some liberties would be willingly foregone.
Out of Rawls’ focus on fairness within social contract theory comes more feminist approaches, like that of Jean Hampton. In addition to criticising Hobbes’ theory, Hampton offers another feminist perspective that focuses on extending the effects of the social contract to interpersonal relationships.
In established states, it can be easy to forget the social contract involves the provision of protection in exchange for us surrendering some freedoms. People can grow accustomed to having their rights protected and forget about the liberty they are required to surrender in exchange.
Whether you think the contract is real or just a useful metaphor, social contract theory offers many unique insights into the way citizens interact with government and each other.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
What do you have a social contract for?
Don’t throw the birth plan out with the birth water!

Don’t throw the birth plan out with the birth water!
ArticleLifestyle + Health
BY Hannah Dahlen 2 AUG 2016
Just try mentioning ‘birth plans’ at a party and see what happens. Hannah Dahlen – a midwife’s perspective
Mia Freedman once wrote about a woman who asked what her plan was for her placenta. Freedman felt birth plans were “most useful when you set them on fire and use them to toast marshmallows”. She labelled people who make these plans as “birthzillas” more interested in birth than having a baby.
In response, Tara Moss argued:
The majority of Australian women choose to birth in hospital and all hospitals do not have the same protocols. It is easy to imagine they would, but they don’t, not from state to state and not even from hospital to hospital in the same city. Even individual health practitioners in the same facility sometimes do not follow the same protocols.
The debate
Why the controversy over a woman and her partner writing down what they would like to have done or not done during their birth? The debate seems not to be about the birth plan itself, but the issue of women taking control and ownership of their births and what happens to their bodies.
Some oppose birth plans on the basis that all experts should be trusted to have the best interests of both mother and baby in mind at all times. Others trust the mother as the person most concerned for her baby and believe women have the right to determine what happens to their bodies during this intimate, individual, and significant life event.
As a midwife of some 26 years, I wish we didn’t need birth plans. I wish our maternity system provided women with continuity of care so by the time a woman gave birth her care provider would fully know and support her well-informed wishes. Unfortunately, most women do not have access to continuity of care. They deal with shift changes, colliding philosophical frameworks, busy maternity units, and varying levels of skill and commitment from staff.
There are so many examples of interventions that are routine in maternity care but lack evidence they are necessary or are outright harmful. These include immediate clamping and cutting of the umbilical cord at birth, episiotomy, continuous electronic foetal monitoring, labouring or giving birth laying down and unnecessary caesareans. Other deeply personal choices such as the use of immersion in water for labour and birth or having a natural birth of the placenta are often not presented as options, or are refused when requested.
The birth plan is a chance to raise and discuss your wishes with your healthcare provider. It provides insight into areas of further conversation before labour begins.
I once had a woman make three birth plans when she found out her baby was in a breech presentation at 36 weeks. One for a vaginal breech birth, one for a cesarean, and one for a normal birth if the baby turned. The baby turned and the first two plans were ditched. But she had been through each scenario and carved out what was important for her.
Bashi Hazard – a legal perspective
Birth plans were introduced in the 1980s by childbirth educators to help women shape their preferences in labour and to communicate with their care providers. Women say preparing birth plans increases their knowledge and ability to make informed choices, empowers them, and promotes their sense of safety during childbirth. Some (including in Australia) report that their carefully laid plans are dismissed, overlooked, or ignored.
There appears to be some confusion about the legal status or standing of birth plans. Neither is reflective of international human rights principles or domestic law. The right to informed consent is a fundamental principle of medical ethics and human rights law and is particularly relevant to the provision of medical treatment. In addition, our common law starts from the premise that every human body is inviolate and cannot be subject to medical treatment without autonomous, informed consent.
Pregnant women are no exception to this human rights principle nor to the common law.
If you start from this legal and human rights premise, the authoritative status of a birth plan is very clear. It is the closest expression of informed consent that a woman can offer her caregiver prior to commencing labour. This is not to say she cannot change her mind but it is the premise from which treatment during labour or birth should begin.
Once you accept that a woman has the right to stipulate the terms of her treatment, the focus turns to any hostility and pushback from care providers to the preferences a woman has the right to assert in relation to her care.
Mothers report their birth plans are criticised or outright rejected on the basis that birth is “unpredictable”. There is no logic in this.
Care providers who understand the significance of the human and legal right to informed consent begin discussing a woman’s options in labour and birth with her as early as the first antenatal visit. These discussions are used to advise, inform, and obtain an understanding of the woman’s preferences in the event of various contingencies. They build the trust needed to allow the care provider to safely and respectfully support the woman through pregnancy and childbirth. Such discussions are the cornerstone of woman-centred maternity healthcare.
Human Rights in Childbirth
Reports received by Human Rights in Childbirth indicate that care provider pushback and hostility towards birth plans occurs most in facilities with fragmented care or where policies are elevated over women’s individual needs. Mothers report their birth plans are criticised or outright rejected on the basis that birth is “unpredictable”. There is no logic in this. If anything, greater planning would facilitate smoother outcomes in the event of unanticipated eventualities.
In truth, it is not the case that these care providers don’t have a birth plan. There is a birth plan – one driven purely by care providers and hospital protocols without discussion with the woman. This offends the legal and human rights of the woman concerned and has been identified as a systemic form of abuse and disrespect in childbirth, and as a subset of violence against women.
It is essential that women discuss and develop a birth plan with their care providers from the very first appointment. This is a precious opportunity to ascertain your care provider’s commitment to recognising and supporting your individual and diverse needs.
Gauge your care provider’s attitude to your questions as well as their responses. Expect to repeat those discussions until you are confident that your preferences will be supported. Be wary of care providers who are dismissive, vague or non-responsive. Most importantly, switch care providers if you have any concerns. The law is on your side. Use it.
Making a birth plan – some practical tips
- Talk it through with your lead care provider. They can discuss your plans and make sure you understand the implications of your choices.
- Make sure your support network know your plan so they can communicate your wishes.
- Attending antenatal classes will help you feel more informed. You’ll discover what is available and the evidence is behind your different options.
- Talk to other women about what has worked well for them, but remember your needs might be different.
- Remember you can change your mind at any point in the labour and birth. What you say is final, regardless of what the plan says.
- Try not to be adversarial in your language – you want people working with you, not against you. End the plan with something like “Thank you so much for helping make our birth special”.
- Stick to the important stuff.
Some tips on the specific content of your birth plan are available here.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY Hannah Dahlen
Hannah Dahlen is a professor of Midwifery at Western Sydney University and a practising midwife.
Join the conversation
How protected are women by the law?
Ethics Explainer: Just War Theory

Ethics Explainer: Just War Theory
ArticleBig Thinkers + Explainers
BY The Ethics Centre 19 JUL 2016
Just war theory is an ethical framework used to determine when it is permissible to go to war. It originated with Catholic moral theologians like Augustine of Hippo and Thomas Aquinas, though it has had a variety of different forms over time.
Today, just war theory is divided into three categories, each with its own set of ethical principles. The categories are jus ad bellum, jus in bello, and jus post bellum. These Latin terms translate roughly as ‘justice towards war’, ‘justice in war’, and ‘justice after war’.
Jus ad bellum
When political leaders are trying to decide whether to go to war or not, just war theory requires them to test their decision by applying several principles:
- Is it for a just cause?
This requires war only be used in response to serious wrongs. The most common example of just cause is self-defence, though coming to the defence of another innocent nation is also seen as a just cause by many (and perhaps the highest cause).
- Is it with the right intention?
This requires that war-time political leaders be solely motivated, at a personal level, by reasons that make a war just. For example, even if war is waged in defence of another innocent country, leaders cannot resort to war because it will assist their re-election campaign.
- Is it from a legitimate authority?
This demands war only be declared by leaders of a recognised political community and with the political requirements of that community.
- Does it have due proportionality?
This requires us to imagine what the world would look like if we either did or didn’t go to war. For a war to be ‘just’ the quality of the peace resulting from war needs to superior to what would have happened if no war had been fought. This also requires we have some probability of success in going to war – otherwise people will suffer and die needlessly.
- Is it the last resort?
This says we should explore all other reasonable options before going to war – negotiation, diplomacy, economic sanctions and so on.
Even if the principles of jus ad bellum are met, there are still ways a war can be unjust.
Jus in bello
These are the ethical principles that govern the way combatants conduct themselves in the ‘theatre of war’.
- Discrimination requires combatants only to attack legitimate targets. Civilians, medics and aid workers, for example, cannot be the deliberate targets of military attack. However, according the principle of double-effect, military attacks that kill some civilians as a side-effect may be permissible if they are both necessary and proportionate.
- Proportionality applies to both jus ad bellum and jus in bello. Jus in bello requires that in a particular operation, combatants do not use force or cause harm that exceeds strategic or ethical benefits. The general idea is that you should use the minimum amount of force necessary to achieve legitimate military aims and objectives.
- No intrinsically unethical means is a debated principle in just war theory. Some theorists believe there are actions which are always unjustified, whether or not they are used against enemy combatants or are proportionate to our goals. Torture, shooting to maim and biological weapons are commonly-used examples.
- ‘Following orders’ is not a defence as the war crime tribunals after the Second World War clearly established. Military personnel may not be legally or ethically excused for following illegal or unethical orders. Every person bearing arms is responsible for their conduct – not just their commanders.
Jus post bello
Once a war is completed, steps are necessary to transition from a state of war to a state of peace. Jus post bello is a new area of just war theory aimed at identifying principles for this period. Some of the principles that have been suggested (though there isn’t much consensus yet) are:
- Status quo ante bellum, a Latin term meaning ‘the way things were before war’ – basically rights, property and borders should be restored to how they were before war broke out. Some suggest this is a problem because those can be the exact conditions which led to war in the first place.
- Punishment for war crimes is a crucial step to re-installing a just system of governance. From political leaders down to combatants, any serious offences on either side of the conflict need to be brought to justice.
- Compensation of victims suggests that, as much as possible, the innocent victims of conflict be compensated for their losses (though some of the harms of war will be almost impossible to adequately compensate, such as the loss of family members).
- Peace treaties need to be fair and just to all parties, including those who are guilty for the war occurring.
Just war theory provides the basis for exercising ‘ethical restraint’ in war. Without restraint, philosopher Michael Ignatieff, argues there is no way to tell the difference between a ‘warrior’ and a ‘barbarian’.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

BY The Ethics Centre
The Ethics Centre is a not-for-profit organisation developing innovative programs, services and experiences, designed to bring ethics to the centre of professional and personal life.
Join the conversation
When are you obliged to go to war?
Learning risk management from Harambe

Learning risk management from Harambe
ArticleLaw + Human Rights
BY Gav Schneider 2 JUN 2016
Traditional and social media channels were flooded with the story of Harambe, a 17-year-old western lowland silverback gorilla shot dead at Cincinnati Zoo on 28 May 2016 after a four-year-old boy crawled through a barrier and fell into his enclosure.
With the benefit of hindsight, forming an opinion is easy. There are already plenty being thrown around regarding the incident and who was to blame for the tragedy.
The need to kill Harambe is exceptionally depressing: a gorilla lost his life, the zoo lost a real asset, a mother was at risk of losing her child, and staff tasked by the zoo to shoot Harambe faced emotional trauma based on the bond they likely formed with him.
In technical risk management terms, the zoo seems to have been in line with best practice.
Overall, it was a bad state of affairs. Though the case gives rise to a number of ethical issues, one way to consider it is as a risk management issue – where it presents us with some important lessons that might prevent similar circumstances from happening in the future, and ensure they are better managed if they do.
In technical risk management terms, the zoo seems to have been in line with best practice.
According to Cincinnati Zoo’s annual report, 1.5 million visitors visited the park in 2014 to 2015. Included in those numbers are hundreds of parents who visited the zoo with children who didn’t end up in any of the animal enclosures.
According to WLWT-TV, this was the first breach at the zoo since its opening in 1978. There is no doubt the zoo identified this risk and managed it with secure (until now) enclosures. There is also little doubt relevant signage and duty of care reminders would have been placed around the zoo. Staff would assume parents would manage their children and keep them safe. In the eyes of most risk management experts, this would seem to be more than sufficient.
Organisations need to put energy and effort into so called ‘black swan’ events… that are unlikely but have immense consequences if they do occur.
However, as we have seen in several cases (including Cecil the Lion) it does not seem to be the incident itself that brings the massive negative consequences but rather social media, based on the fact that the internet provides a platform for everyone to be judge and jury.
This flags the massive shift required in the way we manage risk. If we look only at the financial losses to the zoo, their decision may seem logical and rational. They were truly put in a no-win position – an immediate tactical decision was required once Harambe began dragging the child around the water.
Imagine if they had decided to tranquillise Harambe and the four-year-old boy had died while they were waiting for the tranquillisers to take effect – what would the impact have been?
The true lesson regarding this issue lies in the need for organisations to put energy and effort into so-called ‘black swan’ events – ones that are unlikely but have immense consequences if they do occur. These events are often overlooked, based only on the fact they are unlikely, leaving organisations unprepared for when they do.
The ethics of what is right and wrong tend to blur when the masses have a platform to pass judgement.
Traditional risk management approaches try to allocate scores to things and then put associated resources to the highest ranking risk issues. In this case, a risk that was deemed managed actually occurred and the result was very negative.
Whether negligent or not, various social media commentators have held the mother accountable. It seems she has been held to account based not on what she did, but for the apparently unapologetic and callous way she responded to the killing of Harambe.
This shows us risk management needs to consider the human element in a way we previously haven’t. The ethics of what is right and wrong tend to blur when the masses have a platform to pass judgement. There are many lessons to be taken from this incident, including the following considerations:
- Risk management and duty of care should be incorporated in a more cohesive manner, focusing on applying a BDA approach (Before, During and After).
- Social media backlash adds a new dimension to the way organisations should make, report and defend their decisions.
- Individuals can no longer purely blame the organisation they believe responsible based on negligence or breach of duty of care. Even if the individual shifts blame onto the organisation entirely, and they are not held to account by the law, they will be held to account by the general public.
- We have entered an era where system- and process-based risk management needs to integrate human and emotive elements to account for emotional responses.
- Lastly – and unrelatedly – the question of why one story attracts massive public outcry and why another doesn’t raises ethical questions regarding the way we consume news, the way the media reports it, and the upsides and downsides of social media.
In short this is another case of how much work we still have to do – especially in the modern internet age – to proactively and ethically manage risk.
Ethics in your inbox.
Get the latest inspiration, intelligence, events & more.
By signing up you agree to our privacy policy
MOST POPULAR
ArticleBusiness + Finance
Capitalism is global, but is it ethical?
ArticleBeing Human
It’s easy to ignore the people we can’t see
ArticleScience + Technology
The Ethics of In Vitro Fertilization (IVF)
ArticleBig Thinkers + Explainers
Big Thinker: Peter Singer

Join the conversation
Is the world more just than it used to be?