IB Philosophy 1.4.2 Agency Notes

What Is Agency?

1. Topic 1.4.1 examined self-consciousness and its relationship to personhood. This topic extends the inquiry by asking whether agency, the capacity to act intentionally, is what truly distinguishes persons from other kinds of beings.
2. If personhood requires the ability to choose and act, then the question of who or what possesses agency becomes central to the sub-theme of 1.4.
3. In its broadest sense, agency refers to the capacity of a being to initiate actions based on its own intentions, desires, or reasons. An agent does not merely react to the environment.
4. It acts upon the environment with some degree of purpose. This distinguishes agents from objects such as stones or rivers, which move according to physical causation alone.
5. Aristotle, in De Anima, distinguished between different levels of soul: the nutritive (plants), the sensitive (animals), and the rational (humans). For Aristotle, only beings with a rational soul possess full agency, because only they can deliberate about ends and choose actions on the basis of reason.
6. Aristotle’s three-level hierarchy becomes concrete when we compare: a sunflower tracking the sun (nutritive soul, growth and self-maintenance), a dog learning to sit on command (sensitive soul, perception, desire, locomotion), and a student deliberating over whether to study philosophy or medicine (rational soul, the capacity to reason about abstract possibilities). Only the third, Aristotle argues, constitutes genuine agency.
7. The existentialist tradition places agency at the very centre of human existence. Jean-Paul Sartre argued in Existentialism is a Humanism that human beings are “condemned to be free”: we are agents who must constantly choose and who cannot escape the burden of choice.
8. For Sartre, agency is not an optional feature of personhood, it is the defining condition of being human. Sartre’s famous example of the student illustrates radical agency: a young man during the Occupation must choose between joining the resistance and staying home to care for his ailing mother.
9. No moral theory can decide for him: Kantian duty (the obligation to do what reason demands, regardless of personal cost) points in both directions at once, and utilitarian calculation (weighing who benefits most overall) gives no clear answer either. He is ‘condemned’ to choose freely, and in choosing, he constitutes himself as the kind of person he is.
10. However, the concept of agency raises difficult boundary questions. If agency requires intentional action, what counts as an intention?
11. Simple organisms respond to stimuli in complex ways, and some artificial intelligence systems produce outputs that resemble purposeful behaviour. The question is whether such behaviour constitutes genuine agency or merely its simulation.

Wiredu's Framework and the Broader Implications of Agency

1. Kwasi Wiredu, whose account of personhood as degrees was introduced in 1.4.1, offers a useful framework here. If personhood admits of degrees, then perhaps agency does too: some beings may possess more or less agency depending on their cognitive complexity, social participation, and capacity for moral reasoning.
2. Consider the contrast: a toddler reaching for a hot stove acts on impulse without deliberation. A chimpanzee using a stick to extract termites exercises learned, goal-directed behaviour.
3. An adult deciding to change careers after years of reflection exercises full deliberative agency. Wiredu’s graduated framework captures these differences without drawing a sharp line between persons and non-persons.
4. The IB syllabus asks two pointed questions:
   1. "Could animals or machines be considered persons?" and
   2. "What are the implications of classifying an entity as a person?"
   3. Both questions depend on how we understand agency, because the answer determines which beings deserve moral consideration, legal rights, and social recognition.
5. Agency also connects to the overarching Unit 1 question, "What does it mean to be human?" If agency is uniquely human, then discovering it in animals or machines would challenge the distinctiveness of human existence. If agency is a matter of degree, then the boundary between human and non-human becomes blurred.
6. The relationship between agency and freedom will be explored further in 1.6.1 (Freedom and Determinism). For now, the key point is that agency presupposes some capacity for self-directed action, and the philosophical debate centres on whether this capacity is unique to humans or shared more widely across the natural and artificial world.
7. Understanding agency is therefore essential for evaluating arguments about animal rights, artificial intelligence, and the moral status of various beings. The sections that follow examine each of these debates in detail.

Animal Agency and Personhood

1. The question of whether animals can be considered persons depends largely on whether they possess genuine agency. Many animals display complex behaviours: primates use tools, whales and dolphins communicate through elaborate calls, and crows and ravens solve puzzles they have never encountered before.
2. The philosophical question is whether these behaviours indicate intentional action or merely sophisticated instinct.
3. Peter Singer, in Animal Liberation, argues that the capacity for suffering (sentience), rather than agency or rationality, should be the criterion for moral consideration. On Singer's utilitarian view, any being that can experience pain has interests that must be taken into account, regardless of whether it qualifies as a "person" in the traditional philosophical sense.
4. By contrast, Tom Regan, in The Case for Animal Rights, argues that at least some animals are "subjects-of-a-life": they have beliefs, desires, perceptions, memories, and a sense of their own future. Regan contends that such animals possess inherent value and should be treated as ends in themselves, not merely as means.
5. Research in cognitive ethology (the study of animal cognition) has provided evidence that some animals possess capacities previously thought unique to humans.
6. Great apes pass mirror self-recognition tests, suggesting a form of self-awareness. Elephants mourn their dead.
7. Dolphins use names for one another. These findings challenge the assumption that agency is exclusively human.
8. Wiredu’s framework of degrees of personhood can be applied here: perhaps some animals possess a degree of agency sufficient to warrant a degree of moral consideration, even if they do not reach the full personhood that Wiredu associates with moral achievement (becoming a fully recognised, responsible member of a moral community). This graduated view avoids the all-or-nothing dichotomy between "person" and "non-person."

Kantian Resistance and the Implications of Animal Personhood

1. The Kantian tradition has generally resisted extending personhood to animals. Kant argued that only rational beings are persons, because only rational beings can understand and follow moral laws.
2. On this view, animals may deserve indirect moral consideration (we should not be cruel to them because cruelty corrupts our character), but they are not persons with rights.
3. The implications of classifying animals as persons would be profound. If great apes are persons, then confining them in zoos, using them in medical research, or destroying their habitats would constitute violations of their rights.
4. Several legal efforts (such as the Great Ape Project) have sought to extend basic legal personhood to non-human primates.
5. Buddhist philosophy, with its emphasis on compassion for all sentient beings, offers a perspective that extends moral consideration beyond the human without necessarily granting personhood. In Buddhist thought, all sentient beings participate in the cycle of rebirth, and compassion (karunā) requires that we treat all beings with care, regardless of their cognitive capacities.
6. In the Jātaka tales, the Buddha-to-be in a previous life encounters a starving tigress about to eat her own cubs. He offers his own body as food, demonstrating karunā so deep it extends moral concern to all sentient beings without exception.
7. This challenges any agency framework that restricts moral consideration to rational agents alone.
8. The debate about animal personhood reveals a fundamental tension between two approaches: one that grounds moral status in cognitive capacities (rationality, self-awareness, agency) and one that grounds it in sentience (the capacity to suffer). IB Philosophy students should be able to articulate and evaluate both positions.
9. This question also connects to the discussion of relations to others in 1.2.3: if "the other" includes non-human animals, then our moral community is far wider than traditionally assumed, and the boundaries of personhood must be redrawn accordingly.

Machine Agency and Artificial Personhood

1. The question of whether machines could be considered persons has become increasingly urgent as artificial intelligence systems grow more sophisticated. If agency is the criterion for personhood, and if a machine can act in ways that appear intentional and self-directed, then the possibility of machine personhood cannot be dismissed without argument.
2. Alan Turing, in his famous 1950 paper, proposed what is now called the Turing Test: if a machine's behaviour is indistinguishable from that of a human in conversation, we should attribute intelligence (and by extension, agency) to it. Turing's approach is behaviourist: it focuses on what the machine does, not on what it "is" internally.
3. John Searle challenged this view with his Chinese Room argument. Searle imagines a person in a room who follows rules to manipulate Chinese symbols without understanding Chinese.
4. The person produces correct outputs, but has no understanding of what the symbols mean. Searle concludes that machines that merely process symbols lack genuine intentionality (the property of mental states where they are genuinely "about" something in the world, not just pattern-matching) and therefore lack genuine agency.
5. The functionalist response to Searle holds that what matters for agency is the functional organisation of a system (how its parts work together to produce outputs), not the physical material it is made of. Functionalism says: if a machine's internal processes do the same jobs as the corresponding processes in a human brain, then the machine has genuine agency, regardless of whether it is made of neurons or silicon chips.
6. On this view, silicon-based agency is no less real than carbon-based agency. Consider a thought experiment: a surgeon gradually replaces your neurons one by one with silicon chips that perform identical functions.
7. At each step, your behaviour and reported experiences remain unchanged. If functionalism is correct, you remain a conscious agent throughout, because what matters is the functional organisation, not the material substrate.
8. This raises a profound question: if a machine replicates the right functional structure, must we recognise it as an agent?
9. Modern artificial intelligence systems, particularly large language models and autonomous robots, have reignited this debate. These systems can generate novel responses, adapt to new situations, and in some cases act in ways that their creators did not explicitly programme.

Modern AI Systems and Degrees of Agency

1. The question is whether this constitutes genuine autonomy or merely very complex computation.
2. Consider a self-driving car approaching an unavoidable collision: its algorithm must ‘decide’ whether to swerve left (endangering the passenger) or right (endangering a pedestrian). This scenario forces the agency question into practical urgency: the car exhibits goal-directed behaviour and makes consequential choices, yet lacks anything resembling subjective experience (the felt inner sense of what it is like to be you in that moment).
3. If agency requires felt deliberation, the car is not an agent. If agency is functional, it may be.
4. The implications of granting personhood to machines would be far-reaching. If an AI system is a person, it would possess rights: the right not to be switched off, the right not to be reprogrammed against its will, and possibly the right to legal protection.
5. These consequences force us to think carefully about the criteria for personhood we use.
6. Wiredu’s graduated account of personhood is again relevant. If personhood is a matter of degrees, then perhaps some AI systems possess a minimal degree of functional agency without attaining full personhood.
7. This would allow us to extend some moral consideration to machines without committing to the full range of rights and responsibilities that come with human personhood.
8. The distinction between moral agents and moral patients is useful here. A moral agent can act rightly or wrongly and be held responsible.
9. A moral patient can be harmed or benefited but cannot be held responsible. Some philosophers argue that advanced AI systems might qualify as moral patients (beings we should not harm) even if they are not moral agents (beings who can be morally responsible).
10. This topic connects directly to 9.1.1 (Nature and Role of Technology), which will examine the broader philosophical challenges arising from technological development. It also connects to 1.6.1 (Freedom and Determinism): if a machine's actions are entirely determined by its programming and inputs, can it truly be said to possess freedom and therefore agency?
11. The debates about animal and machine personhood ultimately force us to confront a deeper question: is personhood a natural category that we discover in the world, or is it a social and moral construct that we extend to beings based on our values and interests? How we answer this question has profound implications for ethics, law, and our understanding of what it means to be human.