Tuesday, March 30, 2021

Technology and the Value of Trust: Can we trust technology? Should we?

Can we trust technology? Should we try to make technology, particularly AI, more trustworthy? These are questions that have perplexed philosophers and policy-makers in recent years. The EU’s High Level Expert Group on AI has, for example, recommended that the primary goal of AI policy and law in the EU should be to make the technology more trustworthy. But some philosophers have critiqued this goal as being borderline incoherent. You cannot trust AI, they say, because AI is just a thing. Trust can exist between people and other people, not between people and things.

This is an old debate. Trust is a central value in human society. The fact that I can trust my partner not to betray me is one of the things that makes our relationship workable and meaningful. The fact that I can trust my neighbours not to kill me is one of the things that allows me to sleep at night. Indeed, so implicit is this trust that I rarely think about it. It is one of the background conditions that makes other things in my life possible. Still, it is true that when I think about trust, and when I think about what it is that makes trust valuable, I usually think about trust in my relationships with other people, not my relationships with things.

But would it be so terrible to talk about trust in technology? Should we use some other term instead such as ‘reliable’ or ‘confidence-inspiring’? Or should we, as some blockchain enthusiasts have argued, use technology to create a ‘post-trust’ system of social governance?

I want to offer some quick thoughts on these questions in this article. I will do so in three stages. First, I will briefly review some of the philosophical debates about trust in people and trust in things. Second, I will consider the value of trust, distinguishing between its intrinsic and extrinsic components. Third, I will suggest that it is meaningful to talk about trust in technology, but that the kind of trust we have in technology has a different value to the kind of trust we have in other people. Finally, I will argue that most talk about building ‘trustworthy’ technology is misleading: the goal of most of these policies is to obviate or override the need for trust.

1. Can you trust a thing?

Philosophers like to draw a distinction between trust and mere reliance. The distinction is usually parsed like this: trust is something that exists between people; mere reliance can exist between people and things. One person trusts another when they expect the other will act with goodwill towards them and live up to their obligations. Mere reliance involves the expectation that someone or something will follow a predictable pattern of behaviour.

I believe this distinction was first articulated by Annette Baier in her 1986 article ‘Trust and Antitrust’. More recently, Katherine Hawley has made it the centrepiece of her theory of trust. In her article ‘Trust, Distrust and Commitment’ she opens with a section entitled ‘Trust is not Mere Reliance’. Why not? Hawley accepts that this distinction is not one that is respected in ordinary language. Much to the annoyance of philosophers, people do talk about trusting their cars, their appliances and even the ground on which they walk. But Hawley thinks people are wrong to do so. They should make the distinction because trust is a normatively richer concept than mere reliance. 

Here is the core of her argument:

The distinction is important because trust, not mere reliance, is a significant category for normative assessment. Trust, unlike mere reliance, is connected to betrayal. Moreover trustworthiness is clearly distinguished from mere reliability. Trustworthiness is admirable, something to be aspired to and inculcated in our children: it is a virtue in the everyday sense, and perhaps in the richer sense of virtue ethics too. Mere reliability, however, is not. A reliable person is simply predictable: someone who can be relied upon to lose keys, or succumb to shallow rhetoric, is predictable in these respects, but isn't therefore admirable. Even reliability in more welcome respects need not amount to trustworthiness: when you reliably bring too much lunch, you do not demonstrate trustworthiness, and nor would you demonstrate untrustworthiness if you stopped. 
(Hawley 2012, 2)


This is a strange argument. There seem to be two main parts to it. The first is the claim that trust is linked to betrayal while mere reliability is not. I guess that’s true, but that is probably just an artefact of the conceptual vocabulary we use. Betrayal is the flipside or negative of trust: it’s what happens when trust goes bad. There is, presumably, a negative side to reliability too. Unpredictability? Randomness? The second claim is that trustworthiness is admirable and normatively assessable in a way that mere reliability is not. But is that really true? It seems to me that many people think that being 'reliable' is an admirable quality. I often overhear people talking about work colleagues being reliable with the implication being that they exhibit some virtue. It is true that people can be reliably bad, but that doesn’t say much. After all, people can misplace trust in others or their trust can be betrayed. In other words, just as reliability has its ups and downs so too does trust. I can’t help but wonder if the modifier ‘mere’ is doing a lot of the work in this conceptual distinction. If we said ‘mere trust’ instead of ‘trust’ would we have a similarly dismissive attitude?

In any event, neither of these points is particularly pertinent to the issue at hand. Even if there is this important conceptual distinction between trust and mere reliance, it does not follow that you cannot trust a thing. To make that argument, you would have to suggest that there is some condition of trust that is linked to a property that people have but machines or things lack. What might that be?

The typical answer appeals to mental properties. The idea is that trust depends on having a mind. Since things cannot have minds, they cannot be proper objects of trust. Mark Ryan develops this critique in his article ‘In AI We Trust: Ethics, Artificial Intelligence and Reliability’. In the article, Ryan identifies a number of conditions that must be satisfied in order for trust to exist between two entities or parties. They include things like believing that the other party is competent to perform some action or function, having confidence that they will perform those functions, and being vulnerable to them if they do not. Ryan accepts that machines, specifically AIs, can satisfy these three conditions and so a form of ‘rational’ trust in machines (which might be equivalent to what others call ‘reliance’ or ‘confidence’) is possible. But machines cannot satisfy two other critical conditions for the normatively richer form of trust: (i) they cannot be motivated to act towards us out of a sense of goodwill or out of a desire to live up to their moral obligations toward us; and (ii) they cannot betray us.

Without getting too into the details, I think there are problems with the second part of this argument. Ryan’s claim that machines cannot betray appears to be circular. In essence, his position boils down to the claim that you cannot betray someone unless you are a proper object of trust but you cannot be a proper object of trust unless you have the capacity for betrayal. But that just begs the question: how do you become a proper object of trust or develop the capacity for betrayal? 

That leaves the other part of the argument: the claim that machines cannot have the right kinds of motivation or desire for action. What does Ryan say about this? A lot, but here is one critical quote from his paper:

While we may be able to build AI to receive environmental input and stimuli, to detect appropriate responses, and program it to select an appropriate outcome, this does not mean that it is moved by the trust placed in it. While we may be able to program AI to replicate emotional reactions, it is simply a pre-defined and programmed response without possessing the capacity to feel anything towards the trustor. Artificial agents do not have emotions or psychological attitudes for their motives, but instead act on the criteria inputted within their design or the rules outlined during their development [reference omitted] 
(Ryan 2020, p 13)


In other words, we might create machines that look and act they care about us or look and act like they are motivated by reasons similar to our own, but this is all just an illusion. They don’t feel anything or care about us. They are just programmed artifacts; not conscious, caring humans. They have no minds, no intentions, no inner life.

If you have read any of my previous work on ‘ethical behaviourism’ (e.g. here, here, here and here), you will know that I do not like this kind of argument. To me, it smacks of an unwarranted form of human exceptionalism and mysterianism: humans have this special property that cannot be replicated by machines, but how that property is instantiated in humans is both mysterious and never fully specified. My own view is that while there are important differences between humans and machines (particularly as they are currently designed and operated) there is no ‘in principle’ reason why machines cannot be motivated to act toward us with goodwill and moral rectitude. After all, the only reason we have to believe that other humans are so motivated toward us is because of how they look and act. Looking and acting, broadly defined, are the epistemic hinge on which perceptions of mindedness turn. We can rely on the same evidence when it comes to machines. If they look and act the right way, we can trust them. Similarly, the notion that machines are somehow different from us because they act on the basis of ‘criteria inputted within their design or rules outlined during the development’ also strikes me as being misleading and false. Humans have also been manufactured through a process of evolution by natural selection and personal biological development. We are constrained by both processes and we act on the basis of decision rules and heuristics acquired during these developmental processes. We may sophisticated and complicated biological machines, but there is nothing magical about us.

If I’m right, then even on Ryan’s account of trust it is, in principle, possible for us to trust machines. But this assumes that Ryan (and Hawley and Baier) are right in supposing that trust depends on mental properties like goodwill and a desire to do the right thing. What if that is the wrong way to think about trust?

One of the most interesting recent papers on this topic comes from C. Thi Nguyen. It is called ‘Trust as an Unquestioning Attitude’. In it, Nguyen argues that we can have a normatively rich form of trust in objects and things. Indeed, hearkening back to the point made by Hawley, he suggests that reference to this non-interpersonal or non-agential form of trust are common in everyday language. He cites several examples of this, including climbers who talk about ‘trusting’ their climbing ropes, and people who have lived through earthquakes talking about feeling ‘betrayed’ by the ground beneath their feet.

What is it that unites these non-agential forms of trust? Nguyen argues that this form of trust arises when we have an unquestioning attitude toward something. In other words, when we take it for granted that it will act in a certain way and we depend on it do so. In this respect, we all trust the ground beneath our feet. We don’t wake up in the morning and assume that it will suddenly tear apart and swallow us up. We rely upon this assumption to live our lives. It is only in the extreme case of an earthquake that we realise how much trust we place in the ground. Other examples of this form of trust abound in our everyday discourse.

But what about all those philosophers who insist that trust can only exist between people? Nguyen says something about this:

I have found that philosophers who work on trust and testimony think that this use of “trust” is bizarre and unintuitive — especially locutions like “trusting the ground” and feeling “betrayed by the ground”. But it seems to me that, in fact, these expressions are entirely natural and comprehensible, and it is only excess immersion in modern, narrowed philosophical theories of trust that renders these locutions odd to the ear. 
(Nguyen, MS p 10)


There is a general lesson for philosophers here. For instance, I have encountered a similar phenomenon when writing about gratitude. I once tried to publish a paper on whether atheists could be grateful for being alive. It was repeatedly rejected from journals by reviewers who insisted that gratitude is necessarily interpersonal. According to them, it makes no sense to be grateful for things or for some natural state of affairs. You can only be grateful toward other people. This always struck me as bizarre and counterintuitive but, according to these reviewers, I was the outlier. (If you are interested, you can find the unpublished paper here. Before you say anything, I’m sure there are other reasons why it should have been rejected for publication)

Assume Nguyen is right. What is normatively significant about his version of trust? Nguyen sees trust as an unquestioning attitude as something that is integral to our sense of agency. We are cognitively limited beings. We cannot be constantly suspicious and questioning of everything. By accepting that things (cars, climbing ropes, mobile phones) will work in a certain way, or that people (lovers, friends, fellow citizens) will live up to their obligations, we give ourselves the freedom to live more enriched and open lives.

This doesn’t mean that we can never be suspicious of them. This trust can be misplaced and its wobbly foundations can be revealed in certain circumstances (like in the midst of an earthquake). When this happens we may critically interrogate our previous unquestioning attitude. We may search for data to confirm whether we are right to trust this thing or not. Depending on the outcome of this inquiry, we may find our trust restored or we may find that we can no longer take the thing for granted. Either way, trust as an unquestioning attitude is a normatively essential part of what it means to be human. Given our cognitive limitations, we couldn’t get by without it.

I like Nguyen’s theory of trust. I think it captures something important about our relationship to the world around us. We don’t just rely on our friends, or on the ground beneath our feet or on the smartphones in our pockets. We trust them to act or to persist in certain way so that we can get on with the business of living.

2. Technology and The Value(s) of Trust

If Nguyen’s right, then it does make sense to talk about trust in technology. But this raises a deeper question. Everyone talks about the value of trust but what form does this value take. Is trust valuable in and of itself? In other words, is it a good thing to have trusting relationships in our lives, irrespective of their consequences? Or is trust valuable purely for consequential reasons?

There is a common philosophical distinction that is relevant here: the distinction between intrinsic and instrumental value. It is possible to argue that trust has both kinds of value:

The Intrinsic Value of Trust: Trust is valuable in and of itself (irrespective of its consequences) because it expresses an attitude of respect or tolerance toward the object of trust. For example, if you trust another human being you are signalling to them that you recognise and respect their moral status and moral autonomy.


The Instrumental Value of Trust: Trust is valuable because it is practically essential to human life. It allows us to cooperate and coordinate with others, which allows us to innovate and develop and explore more opportunities. A life without trust would be impoverished because it would lack access to other valuable things.


From my reading of the literature, the instrumental value of trust tends to be emphasised more than the intrinsic value of trust . There is a good reason for this. Everyone that writes about trust notes that trust is a double-edged sword. Whenever you trust a person or a thing you cede some control and power to them. When I trust my partner to look after our daughter, I give up my own attempts to manage and control all aspects of childcare. When I trust my calendar app to keep a record of my appointments and meetings, I give up my own attempts to keep a mental record of my appointments and meetings. The irony is that ceding power and control in this manner can actually be empowering. By not having to worry about childcare or scheduling (at least temporarily) you unlock other opportunities and overcome some of your own cognitive and temporal limitations. This is Nguyen’s argument. But ceding power and control can be risky. The trust can be betrayed. My partner might not look after our daughter properly, my calendar app might fail to update or record a meeting. When this happens I may lose, rather than gain, something that I value.

It is because the consequences of misplaced trust can be so terrible that people tend to emphasise the instrumental value of trust. Even if trust has some intrinsic value this can be swamped by its negative consequences. Imagine if my partner, through neglect, causes our daughter to become seriously ill. By trusting her to look after our daughter I will have expressed my respect for her moral status and autonomy, but that will be of little consolation if our daughter is seriously ill. The intrinsic value of trust is present and cannot be denied, but it has been overridden by the negative instrumental value of trust. It is the instrumental value that matters most.

How is this relevant to the debate about trust in technology? Well, if we accept that we can trust technology (and that it is meaningful to talk about such trust), then we can also accept that this form of trust can have significant instrumental value. It can help us to access other values that would be impossible (or very difficult to obtain) without that trust. But the intrinsic value of trust does seem to be absent when it comes to our relationships with technology. If we accept that most technologies as they currently exist lack an independent moral autonomy and moral status, then we cannot express respect or tolerance for technology by trusting it. This means that the value of trust in technology hinges entirely on the consequences of this trust: if the consequences are good, then it has instrumental value; if the consequences are bad, it does not.

There are three counterarguments to this claim that trust in technology lacks intrinsic value. The first is to claim that even if technology currently lacks independent moral autonomy and status, it may someday acquire this. The typical way to run this counterargument is to suggest that sophisticated machines might acquire the mental properties that we typically associate with moral autonomy and status and, once they do, we will be able to express respect and tolerance toward them by trusting them. Given my earlier critique of Mark Ryan’s views on trust in technology, and my defence of ethical behaviourism, I am quite sympathetic to this argument. I’m just not sure that any present technology rises to requisite level of sophistication.

The second counterargument is to claim that entities do not need to posses mental properties in order for them to have a moral status that is worthy of respect. Environmental ethicists, for example, might argue that aspects of the natural world have an independent moral status that is not derived from human enjoyment of or dependence on the natural world. It is, consequently, not absurd to suggest that we can express respect or tolerance toward aspects of the natural world. If that is right, then it may be less of a stretch to say that trust in technology in its current form has some intrinsic value (remembering, at all times, that this intrinsic value can be swamped by the negative consequences of misplaced trust).

The third counterargument is to claim that technology is a product of human moral agency and autonomy and hence it can have a kind of derived moral status. In other words, it makes sense to express respect for the technology because in doing you are expressing respect for its human creator. There may be some plausibility to this argument in certain contexts. For example, I trust the chef at my favourite restaurant not to poison me. As a result, I don’t test the chemical composition of his food every time it comes out to my table. I just eat it. By trusting that the food will be fine I am, in a sense, expressing my respect for him. But whether this reasoning holds up in the case of technology is much less clear. Most technologies are created by teams of humans. You are not singling any one of them out for respect and, arguably, it is just as mistaken to respect an entire group of humans as it is to respect a thing. But even if you can, the value of trusting their product is still only a derived value and it is quite a nebulous and partial one at that.

In conclusion, trust in technology can have instrumental value (or disvalue as the case may be), but it probably lacks the intrinsic value that arises from trust between human beings. That said, the intrinsic value of trust is quite limited and can easily be swamped by the negative consequences of misplaced trust. So to say that trust in technology lacks intrinsic value is not to say all that much.

3. Concluding Thoughts

None of this is to suggest that we ought to trust technology. It is simply to say that it is meaningful to talk about trusting technology and this type of trust can have significant instrumental value in our lives. Whether it does, in fact, have such value depends on the properties and dynamics of the technology. What does it actually do in our lives? Does it empower us? Or does it act against our interests? Does it do more of the former than the latter?

These are the very same questions we should ask about our relationships with other human beings. We shouldn’t trust all humans. That would be a mistake. Whether we should trust them, or not, depends on who they are and what they do to us. If we take an unquestioning attitude toward them, does this unlock other opportunities and goods for us? Or does it leave us exposed to exploitation and abuse?

It is undoubtedly true that many of us trust technology in our daily lives and are rewarded for doing so. Right now, as I write these words, I’m trusting my computer and my word processing software to safely record and save them for later retrieval. I don’t doubt that the files will be there tomorrow morning when I wish to work on them again. Similarly, I trust my car not to breakdown when I drive to collect my daughter this afternoon. I don’t meticulously check the undercarriage or the engine every time I hop into the driver’s seat.

The problem is that this trust is sometimes betrayed. Modern technologies can let us down. Digital technology is vulnerable to security hacks and data leaks. Mass surveillance can compromise our privacy. Apps can work more for their creator’s interests than for those of their human users. To use a trite example, it is in Facebook’s interests to keep you hooked on their newsfeed and clicking on their ads. Whether this is in your interest is much more doubtful. In many cases, assuming that the technology has a benign effect on your life can be mistaken. This is the dark side of trust.

What can we do about this? Efforts to create trustworthy technology can help, but many of these efforts must be understood for what they really are. Sometimes they are not about encouraging or facilitating trust in technology. They are, instead, about making it possible for us to critically scrutinise the technology. In other words, to make it possible for us to take questioning attitude toward it when we feel unsure about its bona fides. This is why there is such a significant emphasis on transparency, accountability and audit trails when it comes to creating trustworthy technology.

These are laudable goals, and once the mechanisms of accountability are put in place people may well slip back into an unquestioning attitude toward technology. Trust could then be restored. But the policy itself is motivated by the belief that the technology in its current form is not trustworthy.

Friday, March 26, 2021

89 - Is Morality All About Cooperation?

What are the origins and dynamics of human morality? Is morality, at root, an attempt to solve basic problems of cooperation? What implications does this have for the future? In this episode, I chat to Dr Oliver Scott Curry about these questions. We discuss, in particular, his theory of morality as cooperation (MAC). Dr Curry is Research Director for Kindlab, at kindness.org. He is also a Research Affiliate at the School of Anthropology and Museum Ethnography, University of Oxford, and a Research Associate at the Centre for Philosophy of Natural and Social Science, at the London School of Economics. He received his PhD from LSE in 2005. Oliver’s academic research investigates the nature, content and structure of human morality. He tackles such questions as: What is morality? How did morality evolve? What psychological mechanisms underpin moral judgments? How are moral values best measured? And how does morality vary across cultures? To answer these questions, he employs a range of techniques from philosophy, experimental and social psychology and comparative anthropology.

You can download the episode here or listen below. You can also subscribe on Apple PodcastsStitcherSpotify and other podcasting services (the RSS feed is here).


Show Notes

Topics discussed include:

  • The nature of morality
  • The link between human morality and cooperation
  • The seven types of cooperation 
  • How these seven types of cooperation generate distinctive moral norms
  • The evidence for the theory of morality as cooperation
  • Is the theory underinclusive, reductive and universalist? Is that a problem?
  • Is the theory overinclusive? Could it be falsified?
  • Why Morality as Cooperation is better than Moral Foundations Theory
  • The future of cooperation

Relevant links

Thursday, March 18, 2021

The Importance of Collective Intelligence in a Sustainable Future

[This is the text of a short - 15 minute talk - I delivered to the Viridian Conference on 17th March 2021. The purpose of the conference was to discuss the Viridian Declaration, which advocates for technological and social reform in order to make the environment sustainable. One aspect of the declaration focuses on the importance of resilient and adaptive social institutions. That’s where I focused my energies for my talk. As with all short talks of this nature, the ideas and arguments here are programmatic and provocative. They are not rigorously defended. I’m fully aware that there are some holes in what I have to say, but I hope it provides food for thought nonetheless. I have changed the title and made some minor updates to this text from the version I used during the actual talk].

I am very pleased to be able to talk to you this evening and I would like to thank the organisers for inviting me to participate in this exciting and optimistic event. I am a philosopher and ethicist of technology and it is, sadly, rare for me to participate in an event in which people talk positively about the future.

In the short time I have with you I want to make an argument, or maybe even a plea: that we need to think carefully, systematically and scientifically about how best to harness human and machine intelligence to solve the ecological, sociological and technological problems we currently face. Indeed, I would argue that thinking systematically and scientifically about how to harness human and machine intelligence is the greatest challenge of our times. There are some tentative and fascinating attempts to do this already, but they are insufficient. We need to do much more.

To support this argument there are four propositions I wish to defend in the remainder of my talk.

  • Proposition 1: In order for our societies to have a bright and sustainable future, we need to invest in our problem-solving capacity. Since intelligence is equivalent to problem-solving capacity another way of putting this is to say that we must invest in our intelligence capital. This type of capital, not simply labour or technology or natural resources, is crucial to our long-term survival.

Why is this true? I take my support for this from Joseph Tainter’s argument in The Collapse of Complex Societies. For those who do not know, Tainter’s book tries to explain why complex historical civilisations collapsed. He looks at all the famous examples: the Western Chou Empire, The Egyptian Old Kingdom, The Hittite Empire, the Western Roman Empire, and Lowland Classic Maya. After reviewing and dismissing typical explanations for this collapse — including the caim that environmental damage and destruction is the main culprit — Tainter concludes that insufficient of problem-solving capacity is the root cause of civilisational decline. His argument is relatively simple and consists of five claims:

  • Human societies are organisations that solve basic existential and psychological problems for their members. They thereby generate benefits (B) for their members and, in order to sustain themselves, they must continue to solve problems and generate benefits.

  • Like all organisations, human societies must capture and expend energy (C) in order to sustain their problem-solving capacity. Classically, societies have captured energy by foraging, farming, burning fossil fuels, and also through war and imperial expansion.

  • Therefore, in order to sustain themselves, societies face a basic cost-benefit equation: the benefits of increased energy expenditure on problem-solving capacity must exceed the costs (i.e. B must be > C)

  • Increased investment in problem-solving capacity usually yields higher returns (B ↑) but this often comes at an increasing per capita cost so that, at a certain point, the marginal benefits (mB) of increased investment are outweighed by the marginal costs (mC).

  • If mC exceeds the mB at an increasing rate, societies will collapse (i.e. experience a rapid and significant decline in socio-political complexity: this can encompass war, famine, or partition and fragmentation).

The only way that societies avoid this dilemma is by developing some breakthrough technology (broadly understood) that increases the net benefit or reduces the per capita cost of problem-solving. Past examples of such breakthroughs include new methods of food production, facilitation of open markets and trading, new techniques of bureaucratic management, new forms of energy production and, most recently, computing power. But each of these technological breakthroughs is made possible by human intelligence. Hence it is our intelligence capital, not technology or innovation per se, that should be the main target for investment if we are to sustain our civilisation.

  • Proposition 2: The most effective forms of intelligence are collective in nature, not individualistic. In other words, when investing in intelligence capital we should focus primarily on making teams of humans and machines enhance their collective intelligence, and not solely on making individuals more intelligent.

We all like to celebrate lone geniuses such as Einstein, Newton and Darwin. But the reality is that lone geniuses play a limited role in the sustainability and success of human civilisation. It is collections of individuals, coordinating and cooperating their intelligences that makes this possible. There is a certain amount of common sense behind this argument: look around and note how dependent you are on the intelligence of others to solve your basic existential problems. But there have been some books that defend this common sense idea in more detail. One of my favourites is the psychologist Joseph Henrich’s book The Secret of Our Success, which is a systematic defence of the importance of collective intelligence in human history. Henrich uses an arresting thought experiment to set up his main thesis. Imagine you took a random human and a random chimpanzee and dropped them both in the jungle to fend for themselves. Who would you bet on surviving for more than a few days? Henrich argues that you would bet on the chimpanzee every time. Why? Because individual humans are far too dependent on their cultures, their collaborators and their technologies to get by.

Technologies, including artificial intelligence, play a crucial role in collective intelligence structures (or ‘groupminds’ as we might call them). Humans are a technological species. One of the main products of our intelligence is technology: from stone axes to nuclear reactors, groups of humans create and use technologies to enhance our collective intelligence and solve our existential problems. Artificial intelligence is just the latest product of human collective intelligence. Some worry that it threatens to replace or undermine human intelligence. Some of these fears may be well-founded, others less so. That’s not a debate I wish to get into right now. What I do wish to say, however, is that ideally we shouldn’t think of AI as something separate or alien from human collective intelligence. Ideally, we should view AI as a new partner or assistant to human intelligence — as something that needs be incorporated effectively into our collective intelligence structures. Indeed, doing this could give us the technological breakthrough we need to avoid Tainter’s dilemma.

  • Proposition 3: We can think systematically and scientifically about collective intelligence structures and how best to design them to maximise their benefits relative to their costs.

We know that some collective intelligence structures are more successful than others. We also know that there are many different designs or structures that we could implement. We are now starting to develop a scientific field of inquiry into collective intelligence and it is important that we double-down on this effort if we are to survive and thrive in the future.

There are several disparate fields of inquiry that are relevant to this, from organisational theory and management theory, to human-computer interaction studies. At the moment, most of these inquiries are siloed in their own academic subfields and focused on a narrow set of problems. For example, organisational theory and management theory often measures the success of groups in terms of their efficiency or profitability, not their intelligence per se. What we need to do is unite these fields of inquiry into a common discipline centred on collective intelligence.

The work of Thomas Malone and his colleagues has been particularly insightful and innovative in this regard. They have developed group intelligence tests that replicate individual IQ tests. In initial experiments on this idea they find that are ways to measure collective intelligence and that there are some factors that seem to correlate with increased group intelligence. For example, while individual intelligence (as measured by IQ) is one relevant variable, it is not the only one, nor perhaps even the most important one. Indeed, having many high IQ individuals in a group can hinder rather than help collective intelligence. Other crucial variables include cognitive diversity, social perceptiveness of group members (including capacity for empathy and understanding), equality of participation and proportion of women. Malone cautions against overinterpreting these initial findings. There may be more noise than data in them (though the link between cognitive diversity and collective intelligence seems robust) but this is indicative of the kind of inquiry I think we need.

Linked to this, we need to think carefully about the different types of collective intelligence structure and their distinctive costs of benefits. In principle, there are as many structures as there are groups, but Malone argues that we can bring some order to the apparent chaos. He suggests that there are five main types of collective intelligence structure that humans have used to solve our problems in the past:

  • Hierarchies - These are group structures in which a single individual or small group of individuals solves problems for a larger group. In doing so, they are often supported by a legal- bureaucratic agency (or set of agencies).
  • Democracies - These are group structures in which all defined members of the group have a say in group problem-solving, usually through a formal voting procedure. The group decision or output is the aggregated sum of these votes.  
  • Markets  - These are group structures in which individuals are left to themselves and then have to transact and trade with other individuals to solve problems. The group decision or output is then an emergent property of these individual transactions. Markets usually require some external support from a hierarchy or community to ensure that the transactions and trades are enforced.
  • Communities - These are group structures that lack any formal hierarchy or voting procedure and instead involve individuals informally agreeing upon and enforcing some set of social norms for cooperation and coordination. The group solves problems by coordinating on these norms.
  • Ecosystems (state of nature) - These are not group structures but, rather, a form of anarchical social interaction. There are no norms, no voting procedures, no enforceable market transactions. The most powerful individual typically thrives while others scramble for survival.

Malone argues that each of these structures has different benefits and costs for the members of the groups and different ways of distributing the benefits to members of the group. For example, hierarchies often produce significant group benefits (aggregate social surplus) but don’t distribute those benefits evenly among the members of the society. They also tend to be quite costly to maintain: you need create cultural myths/religions, laws and enforcement agencies to maintain the  hierarchy. Democracies are somewhat similar — insofar as they produce significant group benefits and are costly to maintain — but they usually distribute group benefits more evenly. Markets and communities are different again. They can produce group benefits at lower costs (there is less formal infrastructure needed to maintain the group) but can be more unpredictable in how they distribute those benefits. Technology plays a key role in each type of structure too by both reducing the costs of group decision-making and increasing the benefits of working with the group.

Malone’s framework is just a preliminary one but it shows how we might begin to think carefully, systematically and scientifically about collective intelligence. Other work that is useful in this regard would include the series of studies by my colleague Michael Hogan on using collective intelligence in policy-formulation and political decision-making, as well as Geoff Mulgan’s thoughts on collective intelligence in his book Big Mind.

  • Proposition 4: We should be pluralists, not monists when it comes to investment in and design of collective intelligence structures. 

It is tempting to think that we need to find the ‘best’ collective intelligence structure or to assume that one structure is necessarily better than the others (e.g. that democracy is better than all the rest). But we should avoid this temptation. The reality is that modern societies blend different kinds of structure. To stick with Malone’s framework, even in liberal democratic states it is not the case that we only rely on democratic intelligence structures. Our democratic political institutions include elements of legal-bureaucratic hierarchy, may well depend on informal community norms for their success, and in turn function as a support structure for markets. In other words, we cannot have the benefits of democracies without hierarchies and communities, nor can we have the benefits of markets without hierarchies and democracies. The different structures work together and have different strengths and weaknesses.

Furthermore collective intelligence structures can decay, stagnate over time. This may mean they cease to function as effective problem-solving mechanisms. We will need to adapt and redesign them when this happens. Many people argue that this is happening now to mature democratic states. A democracy’s primary tools for increasing its problem-solving capacity and its perceived legitimacy is to enlarge the voting franchise and increase transparency and accountability. But many mature democracies may have hit the point of declining marginal returns on these mechanisms. Political stalemate and polarisation are manifestations of this. How we can redesign democracies to reinvigorate their problem-solving capacity is, I believe, a theme that David Wood will be taking up in the next talk and another major challenge for our time.

My bottom line is this: different kinds of structure are optimal for different kinds of problem. Figuring out which structure is optimal for the set of challenges we now face — ecological, sociological, and technological — is the project we should focus on. We have started to think about this in a systematic and scientific way, but we need to do much more.

Thank you for your attention.

Tuesday, March 16, 2021

The Ethics of Teacher-Student Relationships

When I was starting out in my academic career, I was assigned a senior colleague as a mentor. This is not an unusual practice. The hope is that the senior colleague can provide advice on how to navigate the thickets of academic life. I remember at one of our meetings the topic of teacher-student relationships came up. This colleague told me, in no uncertain terms, that any kind of sexual or romantic relationship with a student (graduate or undergraduate) was inappropriate and should be avoided.

Sound advice, but a little bit ironic for two reasons. First, this particular colleague was in a long-term (and by all accounts happy and well-functioning) relationship with a former graduate student. Second, the thought of entering into such a relationship had never crossed my mind nor had it been a feature of our conversation prior to that point. I believe the only reason it had come up was because I was unsure of how to deal with a student whose mother was dying. To say that the advice was disconnected from the context would seem to be an understatement.

If I were to characterise the relationships I have had with my students over the years I would say that they are, for the most part, extremely distant. To be fair, this my normative baseline when it comes to all relationships. I have very few close friendships and I am, for the most part, reclusive and solitary. That said, I probably take this reclusive attitude to extremes when it comes to students. For example, I try to avoid all social gatherings with students. This includes socialising at university-related events. I don’t like to attend formal dinners or graduation with students, nor do I like to hang around and talk to them after guest lectures or other events (I will, of course, talk to them after my own lectures on course-related topics). When I hear of colleagues going to student balls or taking groups of students out for informal dinner or drinks, perhaps to celebrate the start or end of term, I balk at the idea. I have, very reluctantly, been dragged to such events in the past. I find them unpleasant and awkward. My intention is never to participate in them again. I prefer to deal with students in a purely professional capacity, talking to them solely about course work or academic issues.

I’m not sure why I adopt this style of interaction with my students. Perhaps, in part, it is to avoid any risks associated with conflating different relationship styles. Perhaps, in part, it is due to my own social awkwardness and anxiety. Perhaps, in part, it is due to some misguided belief that you shouldn’t reveal too much of yourself to other people, especially students. Whatever the answer may be, it does prompt the question: what is the preferred form of relating to students? And, more particularly, is it ever appropriate to interact with students as something other than just students?

I’ve read about this topic at various points over the years. Unsurprisingly, most of the literature deals with the ethics of romantic/sexual relationships with students and/or the ethics of teacher-student friendships. Relatively few articles and books focus on what the ideal relationship should be. But maybe it is possible to triangulate on this by considering the various arguments that have been offered against romantic relationships and friendships?

That’s what I will try to do in the following article. I will start by reviewing some basic concepts pertaining to the ethics of relationships and highlighting some pitfalls that plague our reasoning about them. Then I will look at the standard arguments offered against teacher-student romantic relationships (which now tends to represent the consensus view) and the more tentative arguments for and against teacher-student friendship (which are more contested). I will conclude by seeing whether anything can be learned from this inquiry about the preferred way of relating to a student.

1. How to Think About Relationship Ethics

Let’s begin with a few observations about relationships and the ethical norms that may or may not be associated with them. Obviously, humans have many relationships in their lives. Indeed, virtually all repetitive social interactions can be categorised as relationships of some kind. Some philosophers and social scientists believe that it is within these relationships that the human moral conscience is formed. For example, Stephen Darwall has argued that being able to take the second-person perspective (i.e. the perspective of the other party in the social relationship) is key to moral reasoning. Similarly, the developmental and evolutionary biologist Michael Tomasello has argued that being able to understand the duties associated with different social roles is responsible for the evolution of the human moral sense. Finally, though it is less popular these days, Lawrence Kohlberg’s developmental theory of moral reasoning suggests that it is the capacity to see and empathise with the other side of our social relationships that represents the emergence of true moral reasoning in children. I could go on, but I won’t. The point is that social relationships have an important role to play in our moral and ethical reasoning.

There are some ethical rules that apply to all relationships, irrespective of their precise character. For example, you shouldn’t harm someone unless you have good cause. But other moral rules are specific to certain relationships. Lawyers, for example, have a duty of confidentiality to their clients. Doctors too. The problem is that relationships can take many different forms. Think of the relationship between parent and child, doctor and patient, boss and employee, brother and sister, two friends, two lovers and so on. The teacher-student relationship is just one among many. How can we think about the ethical rules that apply to such a diverse array of relationships?

One simple way to think about the ethics of our social relationships is to focus on the purpose or telos of the relationship and to use that to determine what the respective duties of the parties to the relationship might be. Many relationships have a function or goal associated with them. Think about the relationship between a doctor and their patient. The purpose of this relationship is to improve the health of the patient. To do this effectively, the patient has to supply the doctor with all relevant information concerning their health; the doctor then has to be well-informed about the best options for care. This gives rise to respective duties: the duty of honesty for the patient and the duty of competence for the doctor. Both are obviously connected to the goal of the relationship.

That said not all relationships serve single or obvious goals. Some relationships serve multiple goals. Furthermore, thinking about certain relationships in terms of goals can seem contrary to their ethical character. For instance, it seems wrong to suppose that the relationship between friends is goal-oriented. It is no doubt true that friendships serve a purpose: companionship, support, entertainment and so forth. But thinking about them solely in terms of these purposes can seem instrumentalising and dehumanising. If my friends no longer entertain me, am I entitled to abandon them or ignore them? Surely not. Some relationships can be instrumentalised in this way, but not all.

That complication notwithstanding, it seems fair to say that the teacher-student relationship is one that can be thought about in purposive or teleological terms. It does serve a goal, namely: to educate the student (in a broad sense). A first pass at the ethics of teacher-student relationships is to say that the duties of the parties (and the ideal mode of relating between them) flow from that goal. A teacher should not do something that subverts or undermines it, and nor should a student. That said, as everyone points out, there is usually an asymmetry of power between the teacher and student (similar to that between a doctor and a patient) which typically means that the burdens are higher on the teacher than they are on the student. The teacher must do more to ensure that the goals of the relationship are fulfilled.

There are, however, some problems with this initial take on the ethics of teacher-student relationships. I’ll mention three here as they will recur in the analysis given below:

The purpose is vague: To say that teachers should educate their students isn’t to say much since people disagree about what education is really about. Is it about knowledge transfer? Providing credentials? Developing the capacity for critical thought and self-reflection? Producing better citizens for a democracy? Helping students find themselves? Each of these has been proposed as legitimate goal for education over the years and each of them might warrant a different mode of relating to students. Furthermore, some people have, no doubt in a self-serving way, argued that the eroticisation of the teacher-student relationship is part of the educational mission. I’ll return to an example of this below.


Relationships often overlap or nest: Humans often pursue multiple different kinds of relationships with people and often have different relationships types thrust upon them due to social circumstance or necessity. For example, many people are friends with their work colleagues; it is not uncommon for parents to teach their children (not just in homeschooling but in mainstream schools too); and some university professors teach friends or colleagues (because they enroll in their courses). This nesting or overlapping of relationships makes their ethical analysis more complicated. Is it always wrong to pursue different kinds of relationships with people at the same time? Should (can?) one type of relationships be kept isolated from other types?


Relationship analogies are common: Humans often use analogies between relationships to determine the ethical rules that apply to them. We analogise between friendship and intimate partnership, for example, to figure out how we should relate to friends and lovers, respectively. Of course, analogical reasoning is common in human life, but it creates challenges when it comes to the ethics of relationships. If someone thinks a teacher-student relationship is like the relationship between a parent and a child, then they are likely to reach a different conclusion about how they should relate to their students than someone who thinks it is more like the relationship between a boss and an employee. This isn’t a purely hypothetical example either. I have had colleagues in the past tell me that they view the relationship between teacher and student as being much like the relationship between parent and child, and hence had a very particular view of their role within that relationship.*


There are other complications but these will suffice for now. In practice, the overlapping of different relationship types, and how this might bear on the purpose of the teacher-student relationship, is probably the most problematic issue and the one that has generated most debate in the literature on teacher-student relationship. So let’s consider two examples of overlapping relationships teachers can have (and have had) with students: sexual relationships and friendships.

2. The Problem with Sexual Relationships

As mentioned, the ethics of teacher-student sexual relationships has tended to dominate writing in this area. In an interesting article on this, William Deresiewicz points out the image of the feckless, morally corrupt, professor, who sleeps with his (it’s almost invariably a ‘he’) students is probably one of the most common fictional motifs of the 20th century. You couldn’t even begin to list all the examples of it. But we can trace the origins of the motif back much further. It’s right there in Plato and his depiction of Socrates opining about the ethics of teacher-student sexual relationships.

There seems to be good reason for this cultural and intellectual obsession. Teacher-student sexual relationships are a major problem. Recent revelations of rampant sexual harassment and assault of students by well-heeled professors, coupled with institutional misdeeds in covering up these affairs, highlight how rampant it is. In tandem with the #MeToo movement, and the broader societal activism against the sexual mistreatment of women and children, the academy is having to reckon with its history of abuse and misconduct. No wonder people are opining about the ethics of such relationships.

Sexual harassment and assault are not quite the same things as consensual sexual or romantic relationships between two adults. But there is a fuzzy line between these two things in the case of teacher-student interactions. Clearly, there are some ’successful’ romantic relationships that began in this form. As mentioned in the introduction, I have interacted with such couples in the past and my own knowledge of them suggests that they were generally happy and well-functioning (who knows what goes on behind closed doors). But given the nature of teacher-student relationships, there are some very good reasons for thinking that sexual relationships between these parties are always fraught with risk. They are, consequently, best avoided.

There are three obvious reasons for this.

First, the power asymmetry between the parties casts a shadow over any alleged consent to such a relationship. Teachers are the more powerful parties within such relationships, at least within a certain institutional context. They have some knowledge or skill that the student lacks and is supposed to learn from them. Even if the student is highly competent and intelligent in their own right, the default assumption is that this asymmetry exists. Furthermore, the teacher often has power over the future of the student, both in terms of their testing and evaluation, and their access to future opportunities (e.g. through reference writing). It’s a complicated question as to whether this power-asymmetry necessarily undermines any consent that might given to a sexual relationship. But you certainly could argue that there is a lingering, implicit threat inherent in the relationship. Even if the teacher doesn’t say anything, the implication or assumption might be that they could use their power to make life difficult for the student if the student does not consent to the sex.

Even if this shadow doesn’t place the relationship within the realms of illegality or crime, it may, at the very least, place it within the category of what Ann Cahill has called ‘unjust sex’. I covered this idea in a previous article. Cahill derived this category of sex from a series of interviews that Nicola Gavey conducted for her book Just Sex?. Gavey interviewed several women about their sexual experiences. Many of these women agreed that they had consented to some sexual encounters in the past but had felt that they had done so in conditions in which their choices were limited and, in fact, they only had one viable option. Cahill builds on this idea by arguing that in certain contexts, there are less powerful parties whose sexual agency can be hijacked by more powerful parties (Cahill focuses on male-female interactions within patriarchal societies, but I believe it is possible to extend her analysis to all relationships involving power asymmetries). The result of this hijacking can be subtle and insidious. The weaker party may be encouraged to signal consent and approval of what the more powerful party desires in order to accredit it, even though they themselves appear to have limited choices. Cahill’s point is that these cases of unjust sex are not equivalent to rape or sexual assault but, rather, lie in a gray zone between rape and ethically permissible sex. Their moral character is tainted, even if it is not completely reprehensible. It seems to me that this might capture a basic problem with all teacher-student sexual relationships.

Second, there appears to good evidence to suggest that these relationships are often harmful to the weaker party in the long-term. Fredrik Bondestam and Maja Lundqvist recently published a systematic review of the empirical research on the prevalence and consequences of sexual harassment in higher education. They found that it is linked to a number of harmful outcomes for both students and staff, but particularly students. Here is the key paragraph from their study. You can find links to the papers they cite in this paragraph in the original piece:

Exposure to sexual harassment in higher education leads to physical, psychological and professional consequences for individuals. Examples such as irritation, anger, stress, discomfort, feelings of powerlessness and degradation are recurrent in research literature. Evidence-based research confirms more specifically that sexual harassment in higher education can lead to depression (Martin-Storey and August 2016; Selkie et al. 2015), anxiety (Richman et al. 1999; Schneider, Swan, and Fitzgerald 1997), post-traumatic stress disorder (Henning et al. 2017), physical pain (Chan et al. 2008), unwanted pregnancies and sexually transmitted diseases (Philpart et al. 2009), increased alcohol use (Fedina, Holmes, and Backes 2018; McDonald 2012; Selkie et al. 2015), impaired career opportunities (Henning et al. 2017), reduced job motivation (Barling et al. 1996; Chan et al. 2008; Harned et al. 2002), and more. Specific job-related factors often include absence, decreased job satisfaction, engagement and productivity, decreased self-confidence and self-image, and persons giving notice from their jobs (Lapierre, Spector, and Leck 2005; Willness, Steel, and Lee 2007). Even observing or hearing about a colleague’s exposure to sexual harassment can generate ‘bystander stress’ and also cause conflicts in the work team (McDonald 2012; Willness, Steel, and Lee 2007). 
(Bondestam and Lundqvist 2020, 405)

Of course, you may dispute the relevance of this since it deals with sexual harassment (i.e. unwanted sexual attention etc) and not consensual sexual relationships but I will simply reiterate that the line between the two is often blurred. Indeed earlier studies of apparently consensual relationships between staff and students suggests that they can have similar effects, particularly when the relationships breakdown, as many relationships inevitably do. Belinda Blevins-Knabe, for example, in her review of such studies, notes that many female students end up regretting these relationships in the long-term and suffer from anxiety, depression and self-esteem-related issues as a result (Blevins-Knabe 1992, 157). She also notes that the professors involved in such relationships often view them as being problematic and unhelpful too: in one study only 1/6 of those that engaged in such relationships found them to be beneficial (Blevins-Knabe 1992, 157).

There are some qualifications I would like to make to this second argument. First, although I have no doubt that teacher-student sexual relationships lead to the negative outcomes listed above, I would be curious to see how they fare relative to other broken relationships. I imagine (though I have never experienced it myself) that relationship breakdown is stressful and anxiety inducing outside of the academic context, and that it can lead to the negative outcomes listed by Bondestam and Lundqvist. Second, and relatedly, one interesting aspect of some of these studies is the extent to which people retrospectively reevaluate their relationships. It is an old study, but Glaser and Thorpe (1986) suggest that this is a common feature of how students view their former relationships with professors. To what extent are such reevaluations to be credited? Could they be tainted by subsequent events? For example could shifting social norms concerning such relationships (i.e. the fact that people view them as less acceptable than they once were) or the fact that the subsequent career of the student didn’t pan out affect how they perceived and how harmful they are felt to be? I’m sure this happens to some extent. But, even if it does, given that the prevailing cultural wind is against teacher-student sexual relationships, this still provides a reason to avoid them in the interests of harm reduction.

Third, and finally, overlapping a sexual relationship with a teacher-student relationship often undermines the goal of the latter relationship: the pursuit of education. In her recent analysis of the topic, Amia Srinivasan makes much of this argument. She claims that the main problem with these relationships is not the lack of consent but the betrayal of the teaching mission. With typical bluntness, she argues that the goal of teaching is to educate students not to sleep with them. Adding a sexual dimension to the relationship distracts from this goal. One or both of the parties can become more interested in the sex than in sharing knowledge and developing intellectual skills. Sex and intimacy can also undermine teacher impartiality and objectivity, which is crucial to the evaluation and assessment of students, as well as the management of educational activities. Even if teachers claim that the relationship doesn’t harm their professional judgment, it surely must to some extent. Institutional fixes such as anonymous grading and/or reassignment of supervisees can address this problem to some extent but it won’t eliminate it completely. On top of all this, the intimate relationship can affect the burden of care and responsibility within the relationship. Ordinarily, we think of the teacher as the one that carries the heaviest burden: they must care for and nurture the students’ intellectual pursuits. But a sexual relationship can flip this around, particularly in the case of a male professor and female student. As Srinivasan notes, suddenly the student might be expected to care for the professor, not vice versa.

This third argument seems valid to me. But there is a counterargument to it. As noted in the previous section, the purpose of teacher-student relationships is vague. What does it mean to ‘educate’ someone? Could eroticisation be part of education? Deresiewicz, in his article on ‘Love on Campus’, suggests that there is something necessarily erotic about good teaching:

Eros in the true sense is at the heart of the pedagogical relationship, but the professor isn’t the one who falls in love… Love is a flame, and the good teacher raises in students a burning desire for his or her approval and attention, his or her voice and presence, that is erotic in its urgency and intensity. The professor ignites these feelings just by standing in front of a classroom talking about Shakespeare or anthropology or physics, but the fruits of the mind are that sweet, and intellect has the power to call forth new forces in the soul. 
(Deresiewicz 2007)


Deresiewicz goes on to clarify that students shouldn’t mistake this erotic passion for sexuality, and professors shouldn’t take advantage of any potential confusion, but others have made the case without those qualifications. Srinivasan discusses the case of Jane Gallop, a professor accused of sexual harassment by her graduate students in the 1990s. Gallop did not deny the accusations but went on to argue that the sexual dimension of her relationship with the students was a sign of pedagogical success not failure:

At its most intense—and, I would argue, its most productive—the pedagogical relation between teacher and student is, in fact, a “consensual amorous relation.” And if schools decide to prohibit not only sex but “amorous relations” between teacher and student, the “consensual amorous relation” that will be banned from our campuses might just be teaching itself. 
(Gallop, quoted in Srinivasan 2020, 1120)


Even Srinivasan, who is strongly opposed to sexual relationships between teachers and students, concedes that there might be something to this:

Certainly those of us who ended up as professors almost invariably did so because some teacher aroused in us intense feelings of infatuation, desire and want. 
(Srinivasan 2020, 1120)


WTF? I have to say that reading this kind of thing is like reading an alien language. Maybe my experiences are radically different from those of my colleagues, but I have never once had such a passionate feeling for or infatuation with a teacher or professor, nor do I believe that my becoming an academic was the result of some passion being aroused within me by a particular individual. Indeed, I cannot remember a single teacher that has had any a major influence on me. Perhaps I am the outlier.

All that said, my interpretation of these claims about the erotic element of teaching is that they are examples of the fallacy that comes from using analogies to understand the normative character of different relationships. People are analogising too readily between sexual relationships and teacher-student relationships to reach the conclusion that there is something erotic or quasi-sexual about good teaching. I agree that good teaching should stimulate curiosity and passion for a subject or mode of inquiry, but I don’t see this kind of passion as being similar to an erotic or sexual passion. They are quite different.

In any event, the potential injustice, harm and distraction associated with teacher-student sexual relationships seems to provide reason enough to avoid them. They will almost always undermine the ethical character of the relationship, not accentuate it.

I should say that there is one obvious exception to this argument: the case where the sexual relationship pre-dates the teacher-student relationship. It’s possible, particularly at university level, that someone could end up teaching a current or former partner who enrolled in a class or degree programme. I’ve heard of this happening in the past. I think this does create difficulties in practice, and should probably be avoided if at all possible (e.g. by reassigning the student to another lecturer/professor). That said, because the relationship did not arise out of the teacher-student relationship it doesn’t carry quite the same risks when it comes to consent or harm (I suspect!).

What about relationships that post-date the teacher-student relationship? The French president Emmanuel Macron is, famously, married to his former high school teacher. They got married 13 years after they originally met but I believe they had an on-again-off-again relationship from about the time that he was 18. I personally find this strange, but I guess having a relationship with a former student is not as ethically dubious as having one with a current student. That said, my own sense of it is that the amount of time that has elapsed since the end of the teacher-student relationship makes a difference. Getting into a relationship immediately after someone has graduated or left a class seems suspicious to me, but getting into one with someone a decade after your previous interactions seems much less problematic. Personally, I would be concerned about any lingering asymmetries of power or hero worship that might leak into the relationship, but these might not be a factor in some cases.

3. Is there a case for friendships with students?

What about friendships between teachers and students? On the face of it, these would seem to be less ethically problematic than sexual relationships. Friendships don’t raise the same concerns about consent nor do they hold the same potential for harm. Furthermore, I find that many of the people I work with are willing to entertain the idea of being friends with their students. This is particularly true at the graduate/PhD student level. Some people have even suggested to me that it is natural for PhD students to become friends with their supervisors over time. Indeed, it may be one of the hallmarks of a well-functioning supervisor-supervisee relationship.

I have my concerns about this. But a lot of this depends on how we characterise ‘friendships’. There are many competing philosophical definitions of friendship. The most famous and influential of these comes from the work of Aristotle. He distinguished between three kinds of friendship: pleasure friendships (which are about getting enjoyment and entertainment from one another); utility friendships (which are about achieving some goal or purpose with another person’s assistance); and virtue friendships (which are about sharing a commitment to the good with another person, engaging in mutually beneficial and supportive acts, and appreciating the other as a person in their own right, not just a source of pleasure and utility). As you might imagine from these descriptions, Aristotle saw the virtue friendship as the highest ideal of friendship. It was the form of friendship to which we should all aspire.

What significance does this have for teacher-student friendships? Well, it seems plausible to say that teachers can have, and perhaps even should have, utility friendships with their students, provided the utility in question is associated with the goal of education. The student can learn something and, in many cases, so can the teacher. And even if they don’t learn something, they get to hone their skills as an educator. It’s a win-win. Furthermore, as part of that utility friendship, teachers and students probably should be friendly with one another. That is, they should be civil, pleasant, tolerant and so forth. If there is too much resistance and antagonism between them, it will hamper the educational mission.

But can the friendships ever be more than that? Can they ever aspire to something like the Aristotelian ideal? In a thought-provoking article, Amy Shuffleton argues that although such friendships are fraught with risk, there can be merit to them. Shuffleton’s argument is all the more provocative insofar as she focuses not just on friendships between adult students and adult professors at university but, also, on friendships between child students and adults.

Shuffleton accepts that there are two major risks associated with teacher-student friendships. The first is the problem of impartiality: if a teacher is friends with a student it raises concerns about their fairness and impartiality in both assessing and facilitating the education of other students. We encountered this problem in connection with the ethics of sexual relationships. It rears its head here again, albeit without the sexual dimension. Shuffleton argues that this problem actually has two elements to it: the fact of partiality and the perception of partiality. As a matter of fact, many teachers who happen to be friends with their students are not necessarily biased in their favour. Nor, she argues, do students expect such bias. If the friendship is an honest one — and not a Machiavellian one — the student should wish to be treated and assessed fairly. But that doesn’t eliminate the perception of partiality: for all their protestations to the contrary, other people might assume that the teacher is biased in favour of their friends. But Shuffleton points out that many other factors affect the perception of partiality. People might think a teacher is biased to male students or white students or students their share their faith or religious beliefs. Teachers have to work to manage those perceptions and sometimes friendships with students might work to counteract such biases.

The second problem is that the friendship might interfere with or distract from the educational mission. Again, this is similar to the concern raised in relation to sexual relationships but where the distraction takes a different form. Shuffleton offers some interesting responses to this. First, she suggests that teacher-student friendships might support and complement the educational mission in at least some cases, e.g. making students more receptive to learning or preparing them for what it means to be an adult in a democratic society. Second, and more interestingly, she argues that some students might benefit from having adult friends, perhaps because they are excluded by their own peer groups. Children, in particular, can be cruel and prey on any differences or eccentricities. Having an adult that tolerates and appreciates difference could be beneficial to a student. This may involve a form of teacher-student friendship. Shuffleton cites an example from her own life in support of this: a friendship she had with a younger male student while teaching English in Krakow. They did not socialise together, but would talk after class and they bonded over a mutual love of art and photography. This boy’s peers did not seem to share his interests in these things. She thinks there was some value to their friendship.

Shuffleton’s overall point is that we face plural moral demands and obligations. There is a danger that, as teachers, we become too rigid and attached to a certain conception of our role and the moral demands associated with it. In short, Shuffleton’s argument is that we shouldn’t let the moral demands of being a teacher distract from the moral demands of being human.

There is much to commend in Shuffleton’s sensitive and thoughtful account of teacher-student friendships. It does give me some pause and encourage me to reconsider my own distant approach to students. Still, I can’t help but worry about the perception of bias and favouritism that might arise from such friendships. I also think that the suggested benefits of such friendships — toleration, respect and appreciation for difference — can be achieved without slipping into friendship. Indeed, the example Shuffleton gives of the boy she befriended while teaching in Krakow doesn’t really strike me as a true friendship. She was friendly with him without being a true friend. At least, that’s how I see it.

4. Conclusion

So what kind of relationship should a teacher cultivate with their students? I started this article by outlining my own practice in this regard: a relationship of (somewhat extreme) professional distance. Is there any reason to think this is the wrong approach?

Not really. What I have suggested is that it makes sense to think that the ethical character of teacher-student relationships should be determined by the purpose of that relationship: to educate the student (in the broad sense). The problem with this is that this purpose is vague. There are many potential definitions and conceptualisations of what it means to educate someone. But even if this purpose is vague, it seems clear that sexual or intimate relationships between teachers and students are fraught with risk, and tend to undermine the goal of education. Furthermore, even friendship, particularly in its more meaningful forms, creates perceptions of bias and distracts from the educational mission. One can be friendly with students — open, tolerant, respectful — without being their friend.

That said, I would qualify this approach in two respects. First, given that the purpose of education is unclear, and that teachers may not even be able to help students achieve that purpose if it were clear, there is a reason to think that I should focus more attention on the ongoing dynamics of my interactions with students and less on whether those interactions achieve some vaguely specified goal. This is similar to the argument I made about the purpose of parent-child relationships some time ago. Second, taking onboard Shuffleton’s point, we shouldn’t let the demands of teaching detract from the demands of basic human decency.

* One colleague once told me that I should have children because children are like students that you can follow throughout their whole lives. It was such a bizarre analogy that it has stuck in my head ever since.


Monday, March 8, 2021

Understanding Legal Argument (1): The Five Types of Argument

I have been teaching about legal reasoning and legal argumentation for years. When I do so, I try to impress upon students that legal argument is both simple and complex.

It is simple because in every legal case there is, in essence, one basic type of argument at the core of the dispute between the parties. This argument works from a general legal rule to a conclusion about the application of that rule to a set of facts. Philosophers and logicians would say that the basic form of legal argument is a syllogism: a simple three-step argument involving a major premise (a general principle or rule), a minor premise (a claim about a particular case or scenario) and then a conclusion (an application of the general rule to the particular case). 

Here is a simple conditional syllogism:

  • (1) If roses are red, then violets are blue. (Major Premise)
  • (2) Roses are red. (Minor Premise)
  • (3) Therefore, violets are blue. (Conclusion)

My view is that legal arguments take on a similar conditional, syllogistic form. There is a legal rule that stipulates that if certain conditions are met, then certain legal consequences will follow. This is the major premise of legal argument. Then there is a set of facts to which that rule may apply. This is the minor premise of legal argument. When you apply the rule to the facts you get a conclusion.

In abstract form, all legal arguments look like this:

  • (1) If conditions A, B and C are satisfied, then legal consequences X, Y and Z follow. (Major premise: legal rule)
  • (2) Conditions A, B and C are satisfied (or not). (Minor Premise: the facts of the case)
  • (3) Therefore, legal consequences X, Y and Z do (or do not) follow. (Conclusion: legal judgment in the case).

To give a more concrete example, imagine a case involving a potential murder:

  • (1*) If one person causes another person’s death through their actions, and they performed those actions with intent to kill or cause grievous bodily harm, and they had no lawful excuse for those actions, then they are guilty of murder and may be punished accordingly.
  • (2*) Cain caused Abel’s death through his actions and in doing so he intended to kill and acted without lawful excuse.
  • (3*) Therefore, Cain is guilty of murder and may be punished accordingly.

Simple, right? Unfortunately it is not. Although this basic argument is the core of all legal disputes it is not the totality of those disputes. The problem is that legal rules don’t just show up and apply themselves to particular cases. There are lots of potential legal rules that could apply to a given set of facts. And there are lots of qualifications and exceptions to legal rules. You have to argue for the rules themselves and show why a particular rule (or major premise) should apply to a particular case. In addition to this, the facts of the case don’t just establish themselves. They too need to argued for and the law adopts a formalised procedure for establishing facts, at least when a case comes to trial.

In this two-part article, I want to examine some of the complexities of legal argument. I do so first by examining the different kinds of argument you can present in favour of, or against, particular legal rules (i.e. for and against the major premise of legal argument). Understanding these kinds of arguments is the main function of legal education. People who study law at university or in professional schools spend a lot of their time examining all the different ways in which lawyers try to prove that a certain rule should apply to a given set of facts.

Several authors have presented frameworks and taxonomies that try to bring some order to the chaos of arguments for legal rules. I quite like Wilson Huhn’s framework The Five Types of Legal Argument, which not only does a good job of reducing legal argument down to five main forms, but also identifies all the different ways of arguing for or against a legal rule within those five main forms. I’ll try to explain Huhn’s framework, in an abbreviated fashion, in the remainder of this article. I should say, however, that I have modified his framework somewhat over the years and I'm not entirely clear on which bits of it are his and which bits are my own modification. Most of it is his. Some bits are mine (and most of the examples are ones that I use in my teaching and not ones that come from Huhn's book).

1. Argument from Text

For better or worse, law has become a text-based discipline. There are authoritative legal texts — constitutions, statutes, case judgments and so on — that set down legal rules. Consequently, one of the most common forms of legal argument is to identify the case-relevant legal texts and then use them to figure out the relevant rule. This is the first type of legal in Huhn’s framework and perhaps the starting point for most legal arguments.

Here’s a real example. Suppose you punch someone in the face on a night out and they accuse you of assault. You get arrested and you hire a lawyer. You ask them whether you are likely to be found guilty or not. The first thing this lawyer is going to do is to look up the rule governing assault cases like this. In Ireland, this rule is to be found in Section 2 of the Non-Fatal Offences Against the Person Act 1997:

Assault Rule: A person shall be guilty of the offence of assault who, without lawful excuse, intentionally or recklessly: (a) directly or indirectly applies force to or causes an impact on the body of another, or (b) causes another to believe on reasonable grounds that he or she is likely immediately to be subjected to any such force or impact, (c) without the consent of the other. [Text modified slightly from the statute to make the consent condition of the offence a little clearer]


The lawyer will then plug this rule into an argument about your case. Depending on the facts, they may say that you are likely to be found guilty or that you acted with lawful excuse (say, self defence) and so are likely to get off.

But it’s not that straightforward. Text-based legal rules rarely are. The terms within the text need to be interpreted. Their true meaning, within the context of the case at hand, needs to be determined. Sometimes the meaning might be obvious or uncontroversial, but many times it is not. The Irish assault rule, for example, contains a number of vague or uncertain terms. What does it mean to apply force ‘directly or indirectly’? Punching someone in the face is direct force, surely, but what if you throw water in their eyes? Is that indirect force? Come to think of it, what is ‘force’ anyway?

Fortunately, the Irish statute answers some of these questions for us. Later on in Section 2 of the 1997 Act it tells us that force includes “(a) the application of heat, light, electric current, noise or any other form of energy, and (b) application of matter in solid liquid or gaseous form”. So that clears up some confusion. But other doubts remain. What does it mean to ‘recklessly’ apply force? What would a ‘lawful excuse’ be? There are answers to these questions though they are not always clear and it would require significant additional argumentation to confirm the true meaning of the rule in this case.

I won’t belabour the point. As Huhn points out in his discussion, all text-based arguments have to be supported by some kind of textual analysis, i.e. a premise that supports a particular interpretation of the rule. This means that text-based arguments tend to take the following general form:

  • (1) The text setting down the legal rule states ‘If conditions A, B and C are satisfied, then legal consequences X, Y and Z follow”.
  • (2) According to textual analysis T, A really means D, B really means E and C really means F.
  • (3) Therefore, the relevant legal rule is ‘If condition A (meaning D), B (meaning E) and C (with meaning F) are satisfied then legal consequences X, Y and Z follow”.

How do you support a textual analysis? Huhn argues that there are three basic forms of textual analysis in law:

Plain Meaning: The text has an obvious plain meaning. Dictionary definitions support this as does publicly understood meaning. (i.e. it does not require any fancy textual analysis)
Canons of Construction: There is some specific rule of legal interpretation that can clarify the meaning of the rule. There are lots of these so-called ‘canons’ of construction. Examples include expressio unius est exclusio alterius (if you expressly mention one thing you exclude another) or ejusdem generis (when items on a non-exhaustive list are of a particular type then all other members of that list are assumed to be of the same type). You could dedicate years to studying all the different canons of construction and not wrap your head around them.
Intratextual analysis: A later part of the same text clarifies what the true meaning of the rule is. So, to repeat the previous example, a later subsection of the Non-Fatal Offences Against the Person Act 1997 clarifies what is meant by the term ‘force’. Sometimes intratextual analysis is more subtle than this, however. Sometimes it is a case of trying to make an entire legal text coherent and non-contradictory by implying a fuller meaning into an earlier part of the text.


How do you critique a textual analysis? Huhn argues that there are six basic forms of critique, two applying to each of the three forms of textual analysis:

  • Argue that either (a) the text is ambiguous or uncertain and so does not have a plain meaning or (b) the text has a different plain meaning.
  • Argue that either (a) the stated canon of construction does not apply to this type of legal text (some canons of construction apply to statutes or their derivatives and others to other kinds of legal text, e.g. contracts) or (b) there is a rival canon of construction that applies to this text and results in a different meaning.
  • Argue that either (a) there is a conflicting inference to be drawn from the same text; or (b) there is a conflicting inference to be drawn from another text (which must be read with this one, e.g. a statute must be consistent with the text of a constitution since the latter is a superior source of law).


One of the great virtues of Huhn’s framework is his attempt to exhaustively catalogue not only the five main forms of legal argument but also the different ways in which to support or attack those arguments. We’ll see this as we discuss the four remaining types of legal argument.

2. Intent/Purpose-Based Argument

The next type of argument is the intent or purpose-based argument. This is a sub-species of textual argument where, instead of looking at the plain meaning or objective meaning of a legal rule, you focus on the intent or purpose behind the rule. This type of argument is commonly used in contractual interpretation where the court figures out the meaning of a contractual term by appealing back to the intentions of the people who wrote it. It is also used in statutory and constitutional interpretation when lawyers focus on the intent of the legislature when drafting a law or the intent of the ‘framers’ when drafting a constitutional text.

Here’s an example of an intent-based argument. There’s a famous US case called In Re Soper’s Estate. It was decided in Minnesota back in the 1930s. The facts are somewhat unusual. It involved a man called Ira Soper who was married to a woman called Adeline Westphal. They lived in Kentucky. Soper must have been unhappy because he faked his own suicide and ran off to Minnesota. While there, he assumed a new identity (John Young) and married another woman called Gertrude Whitby. Technically, since Soper/Young was not dead, his original marriage to Adeline was still valid. Furthermore, since it was an offence to be married to two women at the same time, this meant that Gertrude was not, technically speaking, his legal wife. This technicality becomes important to the case. 

Young entered into business with another man called Ferdinand Karstens. To protect their interests in the business they signed up to an insurance contract (or rather assurance contract) that ensured lump sum payments to their spouses or next of kin in the event of one of their deaths and ensured that the business would be transferred to the other partner at the same time. Since the terms of this deal could apply to either partner and either of their spouses, it was worded in an abstract way. As follows (this is slightly modified from the original):

Upon the decease of either John W. Young or Ferdinand J. Karstens, the Company shall collect the proceeds of the Insurance Policies upon the life of such deceased Depositor, and shall deliver the stock certificates of the deceased partner to the surviving partner and it shall deliver the proceeds of the insurance on the life of the deceased partner to the wife of the deceased partner if living...


You can probably guess the rest. 

Soper/Young really did die and his second wife Gertrude collected the insurance money. But then his original wife found out that he had faked his earlier ‘death’ and came along claiming that since she was still his wife she should receive the insurance money. The court had to figure out what the phrase ‘wife of the deceased partner’ meant in this context.

The majority judgment in the case decided that the phrase was ambiguous given the circumstances of the case. You had to look beyond the plain meaning of the text to the intentions of Soper/Young when signing the agreement to figure out what it really meant. If you did that, they argued that the meaning was clear: Gertrude (the second ‘wife’) was the intended beneficiary not Adeline. This verdict was disputed by a minority judgment which argued that it flew in the face of the plain meaning of the text. There is some logic to the minority judgment, but it is a complicated linguistic issue. Either way, the majority verdict is still a good example of an intent-based argument being used to justify a particular legal rule — in this case a contractual rule — being applied to a case.

Intent-based arguments have the following general form:

  • (1) The legal text says ‘If A, B, C, then X, Y, Z’
  • (2) The intention/purpose of the person/s that drafted the rule was that A means D, B means E and C means F.
  • (3) Therefore, the applicable rule is ‘If A (meaning D), B (meaning E) and C (meaning F), then X, Y, Z’

To support an intent-based argument you need to introduce evidence of the intent or purpose behind the provision. Huhn suggests that there are four main ways of doing this:

Look to the text itself: sometimes texts explicitly state the intention behind them, e.g. some statutes include long titles or preambles that explicitly or implicitly state the intentions behind them.
Look at changes to the text over time: A history of amendments or revisions to a text might reveal the intent insofar as these amendments suggest a refinement of the text to better approximate the intentions of its drafters.
Look at the history behind the text: The text will have been produced in a certain historical context perhaps in response to a particular challenge or controversy. This might suggest a particular intent. For example, the meaning of the US Constitution is sometimes interpreted in light of the historical purpose of the War of Independence and the separation from the United Kingdom.
Look at commentaries on the text: Commentaries on the text at the time it was drafted or amended (e.g. parliamentary debates about a statute) might reveal intent.


Whenever you move beyond the strict wording of a legal text you are entering troubled waters. There is a view out there that lawyers and judges should concern themselves solely with the strict literal wording of the text. They should not add words that are not there or distort the literal meaning with their own preferences or ideas. Furthermore, the idea that some legal texts have intentions or purposes behind them is problematic since they are often drafted by groups of people that may lack a common intention or they may have been intended to provide timeless abstract principles for a society (this is a common argument made about constitutional texts - If you’re interested I have written a couple of papers about some of the philosophical problems with constitutional interpretation).

As much as people would like lawyers and judges to stick to the texts, the reality is that this practically impossible. Part of the reason for this is that we naturally seek intentions and purposes whenever we engage with the written word. It’s just part and parcel of the social and interpersonal nature of language. Furthermore, language is frequently ambiguous, vague or otherwise uncertain. You have to look beyond the text if you are going to make sense of it. Looking for intentions is a good starting point.

There are many ways to defeat intentional arguments. Huhn identifies four main forms of attack and they can work no matter what form of evidence is being introduced to support the argument:

  • Show that the evidence (of whatever form) suggests a different intent or purpose behind the text.
  • Show that the evidence of intent is not sufficient or is ambiguous or inconclusive.
  • Show that the intent that is evidenced does not count because it did not come from a relevant authority/person (only people with legal authority count when it comes to intent/purpose).
  • Argue that the people who wrote the rule could not have anticipated the current facts and so there is no intent guiding the application of the rule in this case.


3. Precedential Arguments

The common law system is famous for its use of case-based reasoning. Judges decide cases and in doing so they create rules that apply to those cases. Under the system of precedent, subsequent judges in subsequent cases have to follow the same rules if they think their cases are sufficiently similar to the older ones. 

It’s a little more complex than that, of course. The system of precedent recognises a hierarchy between courts. The judgments of superior courts in the same jurisdiction have to be followed but the judgments of inferior courts do not. But even if there is no strict rule stating that previous judgments have to be followed, it is common for judges to look to previous cases for guidance or reassurance when deciding present ones. This is true even when there is an authoritative legal text (such as a statute or constitution) that clearly sets out the rule governing the present case. As we just noted, those texts need to be interpreted and are often ambiguous, vague or otherwise uncertain. Consequently, judges look for guidance from previous cases that interpreted and applied the same rule. Hence, precedential arguments are a core part of the law.

Precedential arguments are a form of analogical argument.* Judges examine the facts of two cases  to determine if they are relevantly similar. If they are relevantly similar, they apply the same rule to the two cases (following the rule from the older/superior court case). If they are not relevantly similar, they might apply a different rule, perhaps one coming from another case, or one that they invent/modify to suit the circumstances.

Here is an example. In the English case of AG Reference (No 6 of 1980), two young people got into an argument in the street. They agreed to settle their dispute by fighting each other. One of them sustained bruising to the face and a bloody nose. They were charged with assault causing harm. The question before the court was whether consent could be a defence to this charge since the two people had agreed to the fight. The court held that consent could not be a defence to a charge of assault causing harm. There were some legitimate exceptions to such a charge, such as legitimately organised sporting events, or certain ceremonial/aesthetic rituals (tattooing, ear-piercing), that might otherwise involve activities that we could classify as assault, but this case did not fall within those exceptions. It was just an ad hoc street fight.

In the later English case of R v Brown, a group of men engaged in private, consensual acts of sexual sado-masochism. They were found out, arrested and charged with assault causing harm. Again, they argued that they should be exempted from this charge because they had consented to the acts. The court disagreed with them. They held that this case was like AG Reference (No 6 of 1980) (and some other similar cases) in that it did not fall within the range of legitimate exceptions to the offence of assault causing harm, and in that consent was not a defence to such a charge.

So, despite some obvious dissimilarities between the cases — one involved a public street fight; the other involved private consensual acts of sexual violence — the majority of judges thought there were relevant structural similarities between the cases that justified the application of the same rule.

Interestingly, there were subsequent cases that seemed similar to R v Brown but were distinguished from it by the courts. For example, in the case of R v Slingsby, a man sexually penetrated the vagina and rectum of his wife with his fist. He was wearing a signet ring at the time and as a result of this his wife was cut, developed septicaemia and died. He was charged with assault causing harm. The courts dismissed the charge on the grounds that this was a private, consensual sexual act that had the indirect consequence (and not the primary aim) of causing harm. This made it very different from R v Brown which involved acts whose primary intention was to cause harm (albeit as part of a sexual kink). Hence it was a legitimate exception to the rule. Similarly, in the case of R v Wilson, a man branded the buttocks of his wife with a hot knife, apparently with her consent. He was also charged with assault causing harm but this was dismissed on the grounds that tattooing and aesthetic adornment of this sort fell within the range of legitimate exceptions to the offence. Furthermore, the interests of marital privacy justified not getting involved. This, again, made the case relevantly dissimilar from R v Brown which did not involve married couples and did not involve tattooing.

You might disagree with this. You might think the cases are more similar than the judges suggest or that they are straining to find structural differences to support bigoted or intolerant views. That’s fine and that’s part of how we go about critiquing analogical arguments. Nevertheless, this sequence of cases provides a good illustration of how precedential/analogical arguments can work.

The general form of precedential arguments is as follows:

  • (1) In case P, the legal rule that applied was ‘If A, B and C, then X, Y and Z’
  • (2) This case is similar to case P in all important respects.
  • (3) Therefore, the rule ‘If A, B and C, then X, Y and Z’ should apply to this case.

There is only one way to defend a precedential argument and that is by following the process of analogical reasoning, i.e. carefully review of the facts of each case, note the relevant similarities, and use this to justify the application of the same rule. Since no two cases are perfectly alike, this is always going to be an imperfect exercise and so analogical arguments are often open to challenge. Huhn suggests that there are six main ways to attack such arguments:

  • Show that the two cases are dissimilar in some important respect.
  • Show that there are two competing lines of authority and so (a) it is unclear what the relevant rule should be or (b) this case is more like the rival authority.
  • Show that the cited opinion from the previous case was not a holding but, rather, obiter dictum. (In common law systems there is this idea that some portions of a previous judgment are legally binding — so-called ratio decidendi and some portions are not — so-called obiter dictum).
  • Show that the cited opinion did not command the majority of the court.
  • Show that the relevant authority has now been overruled or replaced by a new rule and so no longer applies.
  • Show that the relevant authority was incorrectly decided and so should be overruled.

Of course, the relevance of some of these attacks will depend on the system of precedent in which one is operating and the way in which the previous case is being used.

I just want to make one final point about precedential arguments. Because of the way in which they work — building analogies between the fact patterns of two or more cases — this type of argument requires some established view of the facts of the case. You have to agree that the present case has certain features and that the previous case shared those features. If the facts are in dispute or are other than what the judge or lawyer claims them to be, this may block the application of the previous ruling. So this type of argument blends a defence of the major premise of legal argument with an implied defence of the minor premise.

4. Tradition or Custom-Based Arguments

The fourth type of argument is the argument from custom or tradition. This is an unusual one. We know, as a matter of fact, that societies follow rules even in the absence of a formal legal system. Hunter-gatherer bands, for example, have rules that members are expected to follow regarding the sharing of food and the treatment of others. These rules are rarely set down in an authoritative text. They are just habitualised within the society. It is sometimes claimed that the common law system has its origins in such traditions or customs. The early common law judges were not following precedent. There were no precedents to follow. They were, instead, recognising and adopting existing customary rules.

As law has become a more text-based discipline, with formalised procedures for creating and promulgating legal rules, the significance of customary or tradition-based rules has waned. Instead of pointing to customs, we point to texts to find the rules that govern our cases. Still, custom is an integral part of the law in certain areas. In contract law, for example, it is common to use customs within certain professions or locales to work out what the terms of a contract ought to be. Similarly, in international law, the customary behaviours of states toward one another is one of the primary sources of law. Finally, and perhaps most notoriously, there is no written constitution in the UK. There is, instead, a set of customary rules and norms that dictate how the state should be run.  These are the main source of constitutional law in the UK.

Some philosophers and political scientists would go further than this. They would argue that since it is impossible to authoritatively and comprehensively write down every rule that governs society, the law must be supported by a significant body of unwritten, tacit, traditionary and customary rules. At some point in time, judges and lawyers must appeal to these rules in order to make legal arguments.

Here is an example of a custom-based argument in contract law. The case in question is the Irish case of Carroll v Dublin Bus. This involved a bus driver who was out of work for a period of time due to illness. He returned to work and was given an ordinary bus route. He disputed this on the grounds that it was a custom/tradition within Dublin Bus that drivers returning to work after a period of illness be given a ‘light’ or ‘rehabilitation’ route before being eased back into a normal work routine. The court agreed that this was indeed a custom within Dublin Bus and so he was successful. The court decided that this custom should be implied into the terms of his employment contract.

Custom or tradition based arguments take the following form:

  • (1) The tradition/custom states that ‘If A, B and C, then X, Y and Z’
  • (2) Evidence shows that the habits and customs of people in a given area/profession (etc) support the traditional rule ‘If A, B and C, then X, Y, and Z’
  • (3) Therefore, the rule ‘If A, B and C, then X, Y and Z’ should apply to this case.

The key to a good custom-based argument is twofold: (i) show that the written law does not cover the facts that arise in the case or is necessarily incomplete without the custom; and (ii) provide evidence to show that people really are committed to that custom. There are a few ways of doing (ii). Huhn mentions the following:

Historical analysis: Show that the historical record supports the idea that people have always followed or endorsed this rule.
Recorded Opinion/ Commentary: Show that the evidence on public opinion (or the opinions of relevant sub-groups of the public) suggest that they agree to this rule.


A combination of methods can be the most effective.

There are three ways to attack a custom-based argument:

  • Show that the alleged tradition does not exist, i.e. the evidence for the tradition is weak or incomplete or unpersuasive.
  • Show that there have been competing traditions and so (a) it’s not clear which traditional rule should apply to this case or (b) the alternative traditional rule should apply to this case.
  • Show that a new tradition is emerging which displaces the old traditional rule (this is a frequent problem with tradition-based argument since society is always changing and adapting to new realities).

5. Policy-Based Arguments

The final type of argument (according to Huhn’s framework) is the policy-based argument. This is perhaps the most contentious type of argument. It involves advocating for the application of certain rules on the grounds that they are good policy, or, conversely, arguing against the application of certain rules on the grounds that they are bad policy. This type of argument is controversial because some people think that lawyers and judges should not be engaged in policy-making, but the practical reality is that policy-based arguments are widespread in law, and they are often critical in the most contentious cases.

Policy-based arguments have two key steps to them. The first is an inquiry into the likely consequences or outcomes of applying a particular rule to the facts of the case (so, again, there tends to be some initial agreement on the facts though it is not as integral to this type of argument as it is to a precedential argument). The second is the use of some evaluative or normative theory to assess those consequences or outcomes. This evaluative theory can be drawn from multiple sources: economic theory, moral theory, and religious tradition are among the most commonly used.

In my experience, it is rare for courts to use policy-based arguments to simply create entirely new legal rules. Maybe that did happen back in the day. What’s more likely to happen nowadays is that there is some dispute as to which rule (or which interpretation of a rule) should apply to a case. To resolve this dispute, courts will examine the likely outcome of applying the rule to the case. If they think the outcome is consistent with their preferred evaluative theory, they will apply it. If not, they will look for an alternative rule (or an alternative interpretation of the rule).

Here’s an example. The English case of Re A (Conjoined Twins) is remarkable for a number of reasons. The facts are well-known. A pair of conjoined twins (referred to as Jodie and Mary in the case) were born in August 2000. Jodie was the stronger of the two. Mary was only kept alive by a common artery that she shared with Jodie. If left conjoined, they would both, almost certainly, die. If separated, Jodie would live and Mary would die. The doctors wanted to separate them. The parents objected. The case was referred to the courts for guidance as to whether it was legally permissible for the doctors to proceed with the separation.

There were many issues in the case. The chief one was whether the surgeons would be guilty of murder if they performed the surgery. Though there were disagreements among the judges, they agreed that the surgeons would be intentionally killing Mary by performing the separation (they would be intentionally causing her death through their actions) but that they did have a lawful excuse for doing so. What that lawful excuse was ended up being disputed between the judges. One judge analysed the case partly in terms of self-defence, but two of them agreed that the doctors could avail of the defence of necessity. In brief, they argued that what the doctors were doing was necessary, in the circumstances, to prevent the death of both twins and save the life of Jodie (it was the lesser evil).

The problem with this, however, is that previous case law suggested that the defence of necessity was not available to a charge of murder. One of the reasons for this limitation, suggested by commentators on the law, was that if it was a defence people might be too quick to appeal to it to sanction their own murderous acts. In other words, we might have an army of would-be murderers suddenly concocting situations of necessity in order to get away with murder. One of the judges in the case, Lord Justice Brooke, argued that this predicted outcome was unlikely to occur:

If a sacrificial separation operation on conjoined twins were to be permitted in circumstances like these, there need be no room for the concern felt by Sir James Stephen that people would be too ready to avail themselves of exceptions to the law which they might suppose to apply to their cases (at the risk of other people's lives). Such an operation is, and is always likely to be, an exceptionally rare event, and because the medical literature shows that it is an operation to be avoided at all costs in the neonatal stage, there will be in practically every case the opportunity for the doctors to place the relevant facts before a court for approval (or otherwise) before the operation is attempted.


This is an example of a policy-based argument. Brooke LJ is considering the possible outcome of applying the defence of necessity to this case and is arguing that it is unlikely to have negative consequences (it’s not going to give an easy excuse to would-be murderers). Hence, he is willing to apply the defence to this case. (Note: there are many other policy-based arguments in the judgment — this is just one example).

Policy-based arguments have the following general form:

  • (1) The supposition/working hypothesis is that rule R (taken from text, intention, precedent or tradition) applies to this case.
  • (2) If rule R applies to this case, good/bad consequence/outcome/moral fit X, Y, and Z will occur. (Prediction Premise)
  • (3) We should adopt a rule with good consequences/outcomes/moral fit; we should not adopt a rule with bad consequences/outcomes/moral fit (Normative Premise)
  • (4) Therefore, rule R should/should not apply to this case.

The key to defending a policy-based argument is to show (a) that the predicted consequence/outcome is likely to occur and (b) that it is consistent/inconsistent with the preferred evaluative theory. There are many different evaluative theories so there are many ways of trying to defend the normative premise of this argument. Still, in broad outline, we can say that there are two basic methods of evaluation. First, there is deontological evaluation where you check to see if the proposed legal rule is consistent with another rule drawn from your preferred evaluative theory (e.g. a secular moral theory such as Kantianism or a religious moral theory). Second, there is pure consequentialist evaluation where you check to see whether the consequences of the rule are good or bad according to the criteria of your preferred evaluative theory (does it promote economic growth? limit suffering? promote welfare and well-being? reduce crime? and so on)

Because of their contentious nature, and in particular because of the widespread disagreement about preferred evaluative theories, policy-based arguments are frequently attacked. Huhn suggests that there are six main methods of attack:

  • Argue that it is not the job of the law to make these policy judgments (that’s a job for the legislature or the public).

  • Show that the relevant evaluative theory actually supports an alternative rule.
  • Show that although the policy is good, it is not served in this case (i.e. the prediction is false).
  • Show that there is a competing policy outcome that should be preferred.
  • Show that the alleged desirable/undesirable consequences will not follow from the rule.
  • Show that policy considerations are not sufficiently strong to outweigh other legal arguments.

6. Conclusion

So there you have it. This is a brief overview of the five main ways of defending and critically analysing the first premise in any legal argument. I've summarised all the main ideas into a handy chart/table which you can download here, if you like.

* There are some legal theorists that claim that this is wrong. But in my experience they offer highly technical analyses of case-based reasoning that are divorced from how legal practitioners actually think it works in practice. It’s simpler, and in my view more accurate, to see precedential arguments as a species of analogical argument.