Intention is the mental commitment to carry out a specific action or achieve a particular outcome. Unlike a passing wish or idle desire, an intention connects a goal to a plan you expect to act on. That distinction sounds simple, but it ramifies into nearly every domain of human life, from whether a crime is classified as murder or manslaughter, to how five-month-old infants make sense of the people around them, to how engineers build robots that can explain their own decisions. The concept sits at the intersection of philosophy, neuroscience, law, psychology, and artificial intelligence, and each field has sharpened the definition in its own way.
How Intention Differs from a Desire or a Goal
People often use “intention,” “desire,” and “goal” interchangeably, but researchers who study motivated behavior draw sharp lines between them. A desire is something you want; an intention is something you have committed to doing. You can desire to run a marathon without ever intending to train for one. The difference matters because desires can be contradictory and still coexist comfortably in your head, while intentions push you toward action and resist contradiction. Research identifies three criteria that separate the two: perceived performability (you believe you can actually do it), action-connectedness (it is linked to concrete steps), and temporal framing (it is anchored to a specific time or context).1European Journal of Social Psychology. The distinction between desires and intentions
A goal, meanwhile, is the endpoint you are aiming for. An intention is the bridge between the goal and the first real step toward it. You might have a goal of saving money, a desire to stop spending on takeout, and an intention to pack lunch every weekday. The goal is the destination, the desire is the emotional pull, and the intention is the commitment that connects the two to behavior. This layered structure is why psychologists and behavioral scientists treat intention as the strongest single predictor of whether someone will actually do something, though it is far from a guarantee.
The Neuroscience of Intending to Act
One of the most famous experiments in modern neuroscience asked a deceptively simple question: when you decide to move your hand, does your brain start preparing before you feel the decision? The answer, first reported in 1983 by Benjamin Libet and colleagues, was yes. Electrical recordings from the scalp showed that the brain’s “readiness potential,” a buildup of activity in motor areas, began at least several hundred milliseconds before subjects reported the conscious experience of wanting to move.2PubMed. Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential). The unconscious initiation of a freely voluntary act This finding sparked decades of debate about free will: if your brain is gearing up before you feel the intention, who is really in charge?
More recent work has complicated the picture in ways that are less headline-friendly but more scientifically honest. Studies have shown that the readiness potential and the moment of felt intention do not actually track each other. They do not rise and fall together across trials, which undercuts the assumption that one causes the other.3PubMed. (Non)-experiencing the intention to move: On the comparisons between the Readiness Potential onset and Libet’s W-time The relationship between brain activity and the feeling of intending to act is more tangled than the original experiment suggested.
Brain-imaging and clinical studies have also revealed that intention is not a single event in one brain region. Research points to at least two distinct stages of motor intention. The first, a sense of “wanting to move,” appears to be generated in the inferior parietal lobule, a region involved in spatial awareness and planning. This early stage specifies a general goal before any detailed movement plan exists. The second stage, an “urge to move,” arises in areas closer to the motor cortex and signals that a planned movement is about to launch.4PubMed. Conscious motor intention emerges in the inferior parietal lobule Intention, at the neural level, is a process that unfolds over time rather than a single switch being flipped.
What Happens When the Intention System Breaks Down
Clinical neurology offers some of the most striking illustrations of what intention actually does by showing what goes wrong without it. In a condition called utilization behavior, patients reach out and use objects placed in front of them even when doing so makes no sense. A patient might pick up a comb and start combing their hair mid-conversation, or put on a pair of glasses that belong to someone else. The problem is not that they cannot move correctly. The problem is that their actions are no longer governed by personal goals. They lose the ability to detect when what they are doing conflicts with what they intended to do. The actions feel normal to the patient because the internal reference point, the intention against which actions are normally checked, is missing or inaccessible. External cues from the environment hijack behavior in the absence of top-down intentional control.
With repetition, even healthy people experience a milder version of this shift. Goal-directed actions, the kind guided by active intention, can become habitual through routine. Once a behavior is habitual, it runs on autopilot and is no longer sensitive to the original goal in the same way. You drive your usual route to work even when you meant to stop at the store. The intention was there, but the habit overrode it. This everyday lapse hints at the same underlying architecture that goes dramatically wrong in utilization behavior: intention and automatic motor responses are separate systems, and they can come apart.
How Intention Develops in Infancy
Humans do not arrive in the world understanding intention, but they get there remarkably fast. By about five months of age, infants already interpret other people’s actions as goal-directed. In a classic experimental setup, babies watch an actor reach for an object. When the object is moved and the actor reaches for it in the new location, infants are not surprised, because the goal (getting the object) stayed the same. But when the actor reaches to the old location even though the object has moved, babies stare longer, as if the scene does not make sense. They are tracking the relationship between the person and the goal, not just the physical trajectory of the arm.5PubMed Central. The emergence of intention attribution in infancy
This early capacity for reading goal-directedness is thought to be a building block for more sophisticated social reasoning that comes later, including the ability to understand that other people have beliefs, knowledge, and plans that differ from your own. Research on primates suggests that this basic form of intention-reading is not unique to humans. When macaque monkeys were familiarized with an actor reaching over a barrier to grab an object, they showed surprise (longer looking times) if the actor continued to use the arcing motion after the barrier was removed, suggesting they expected the actor to take the direct path once the obstacle was gone.6PubMed Central. Early Intention Understandings that are Common to Primates Predict Children’s Later Theory of Mind The monkeys, like human infants, seemed to interpret the action in terms of its goal rather than its surface features.
Do Other Animals Understand Intention?
The primate research goes further than just looking-time experiments. Chimpanzees have been shown to attribute goals to other chimpanzees and to adjust their helping behavior accordingly. In one study, chimps were given a set of seven objects and could choose which one to pass to a partner who was trying to solve a tool-use task. When they could see their partner’s situation, they selected the right tool for the job. When their view was blocked, they still attempted to help but failed to pick the correct tool, because they could not assess what the partner needed.7PubMed Central. Chimpanzees’ flexible targeted helping based on an understanding of conspecifics’ goals This flexible, context-sensitive helping implies that chimpanzees understand something about what another individual is trying to accomplish and can tailor their response to match.
Chimpanzees also appear to attribute goals in simpler scenarios modeled on infant studies, responding much the way human babies do when an actor’s goal is disrupted or redirected.8PubMed. Goal attribution in chimpanzees The capacity is not distinctively human. What may be distinctively human, though, is what we do with it. Humans take this basic ability and layer on top of it a rich understanding of beliefs, plans, and shared commitments, the kind of reasoning that allows us to coordinate large-scale cooperation, build institutions, and hold each other morally accountable based on what someone meant to do rather than merely what happened.
Intention in Criminal Law
Law treats intention as one of the most important factors in determining blame. In criminal law, the mental state at the time of an act, often called mens rea, can mean the difference between a minor offense and the most serious charge available. Two major forms of legal intention are recognized. Direct intention is straightforward: you desired a particular outcome and acted to bring it about. Oblique intention is subtler. It applies when you did not desire the outcome but appreciated that it was virtually certain to result from your actions. A jury may find that a person intended a consequence even if they did not want it, so long as they recognized it as a near-certainty.9Oxford Law Trove. Criminal Law Directions
This distinction creates real controversies. One of the most enduring is the doctrine of double effect, which holds that there is a morally significant difference between intending a harmful outcome and merely foreseeing that a harmful outcome will result from your actions.10PubMed Central. How the Doctrine of Double Effect Rhetoric Harms Patients Seeking Voluntary Assisted Dying A physician who administers high-dose pain relief knowing it may hasten death is, under this doctrine, in a different moral category from a physician who administers the same dose with the purpose of hastening death. Critics argue that when the outcome is identical and foreseeable, the distinction can become a legalistic fiction. Supporters counter that intention is what separates purposeful killing from the acceptance of a tragic side effect, and that collapsing the two would make it impossible to provide adequate palliative care. The debate is unresolved and remains active in bioethics and end-of-life policy.
Culture Shapes How Much Intention Matters
Western legal and moral systems place enormous weight on what a person meant to do. An accidental death is tragic; a deliberate one is monstrous. But this emphasis on intention is not universal. Research across societies shows that the relative importance of intention versus outcome in moral judgment varies dramatically depending on cultural norms around interpreting other people’s mental states.
A cross-cultural study compared moral reasoning among Indigenous iTaukei Fijians from Yasawa Island, Indo-Fijians, and North Americans. North Americans weighed the perpetrator’s mental state heavily: a failed attempt to harm was judged more harshly than an accident that caused equal harm. Yasawan participants showed the opposite pattern, judging accidents more harshly than failed attempts, placing more emphasis on what actually happened than on what someone meant to do. Critically, when Yasawan participants were prompted to think about the actor’s thoughts before making their judgment, their responses shifted toward the North American pattern, suggesting that the difference is not about an inability to reason about intention but about cultural norms that discourage doing so in everyday social life.11PubMed. Weighing outcome vs. intent across societies: How cultural models of mind shape moral reasoning
The finding is a useful corrective for anyone tempted to treat intention-focused moral reasoning as the natural human default. It is one available mode of moral thinking, and cultures differ in how much they use it. Some societies prioritize relationships, consequences, and social harmony over the internal mental states of individuals. Intention still matters in those contexts, but it occupies a less central position in how people assign blame and organize their moral worlds.
From Intention to Action and the Gap Between Them
Anyone who has made a New Year’s resolution knows that having an intention does not guarantee following through. This gap between intending to do something and actually doing it is one of the most studied problems in behavioral psychology. People regularly intend to exercise, eat better, save money, or study more, and then fail to act on those intentions. The gap is not about dishonesty; people genuinely mean to follow through when they form the intention. The problem is that intention alone does not automatically produce the right behavior at the right moment.
One strategy that has shown consistent results in narrowing this gap is the formation of “implementation intentions,” which are if-then plans that specify exactly when, where, and how you will carry out the intended behavior. Rather than intending to “exercise more,” you commit to “go for a run at 7 a.m. on Tuesday and Thursday mornings in the park.” Research has found that forming these specific plans improves goal attainment not only by helping people get started but also by protecting goal pursuit from distractions and unwanted competing thoughts.12PubMed. Implementation intentions and shielding goal striving from unwanted thoughts and feelings The if-then structure essentially delegates part of the job to the environment: when the cue appears, the planned response fires more or less automatically, reducing the need for effortful self-control in the moment.
Intention in Philosophy and the Problem of Future Plans
Philosophers have debated the nature of intention for centuries, but a particularly productive thread in contemporary philosophy concerns future-directed intentions, the kind you form now about what you will do later. These are different from the intention you have at the moment of acting. When you book a flight for next month, you have formed an intention about a future action. That intention now shapes what else you plan, what options you consider, and what you treat as settled. It constrains your reasoning going forward.
The philosopher Michael Bratman argued that future-directed intentions play a fundamental role in making us rational agents over time. They are not just predictions about what we will do; they are commitments that structure our practical thinking, letting us coordinate our own actions across days, weeks, and years. Without them, we would have to start every decision from scratch, unable to build plans that depend on earlier commitments. This view has been influential enough that an entire branch of philosophical work examines whether a theory of rational action can function with or without granting future-directed intentions a special status.13Rational Powers in Action. Future-Directed Intentions and the Theory of Instrumental Rationality
The philosophical discussion connects neatly to everyday life. When you tell a friend “I’ll be there Saturday,” you have not just made a prediction. You have committed to a course of action that both of you now rely on. Your friend makes plans around your stated intention. You feel a pull to follow through not because of some abstract principle but because future-directed intentions create a web of social and practical coordination that would fall apart if people treated their own plans as merely tentative forecasts.
Shared Intention and Collective Action
Much of what humans accomplish depends not on individual intention but on shared intention, the state in which multiple people commit to doing something together. Building a house, playing in an orchestra, organizing a protest: these require more than each person independently intending to do their own part. There needs to be a mutual understanding that you are cooperating toward a shared goal. Philosophers who study collective intentionality debate whether individual intentions to do your part are enough to explain joint action, or whether something additional is required, perhaps a special type of “we-intention” or a complex web of interlocking commitments that all participants recognize and rely on.
This capacity for shared intentionality may be one of the traits that most distinguishes human cognition from that of other primates. While chimpanzees can read individual goals and help accordingly, humans can form and maintain shared plans that extend across large groups and long time horizons. Some researchers have proposed that this ability is what drives cumulative culture, the process by which human societies build on previous innovations in a way that ratchets up over time, allowing each generation to start from a higher platform than the last. Without shared intention, the kind of coordinated teaching, learning, and cooperative problem-solving that makes cultural accumulation possible might never get off the ground.
Intention in Artificial Intelligence
Computer scientists borrowed the concept of intention from philosophy decades ago, and it remains central to how intelligent agents are designed. The Belief-Desire-Intention (BDI) model gives software agents three mental states: beliefs about the world, desires about how they want the world to be, and intentions that commit them to specific courses of action.14Journal of Web Semantics. The Belief-Desire-Intention Ontology for modelling mental reality and agency The BDI framework has been used in everything from air traffic control systems to Mars rover planning. Its strength is that an agent with intentions can explain why it did what it did: “I believed X, I desired Y, so I intended to do Z.” That transparency is valuable when humans need to trust and verify an autonomous system’s decisions.
Recent work has combined BDI agents with large language models to get the best of both worlds. The language model handles understanding natural-language commands from human users, while the BDI agent manages the actual reasoning, goal management, and safety verification. The BDI layer ensures that the robot’s decisions are traceable and verifiable, while the language model makes the interaction feel natural rather than rigid.15Engineering Applications of Artificial Intelligence. Integrating Belief-Desire-Intention agents with large language models for reliable human–robot interaction and explainable Artificial Intelligence
On the user-facing side of AI, the word “intention” shows up in a different sense. When you type a question into a chatbot or virtual assistant, the system needs to classify your intent: are you asking for directions, making a complaint, requesting a refund, or just chatting? Intent classification is a core task in building dialog systems, and it turns out to be harder than it sounds. Systems trained on a fixed set of supported intents struggle when users say something that falls outside the expected categories. A benchmark dataset covering 150 intent classes across ten domains found that while classifiers performed well on queries that matched a known intent, they had a much harder time recognizing when a query was out of scope entirely.16ACL Anthology. An Evaluation Dataset for Intent Classification and Out-of-Scope Prediction Knowing what someone intends, even in the narrow sense of what they are trying to accomplish with a typed sentence, remains a genuine engineering challenge.
Why Intention Resists a Single Definition
Part of what makes “intention” hard to pin down is that it legitimately refers to different things depending on context. The neuroscientist studying motor intention is interested in the brain processes that occur in the fraction of a second before you move your finger. The philosopher studying future-directed intention cares about the commitment you make today about what you will do next year. The criminal lawyer is interested in whether you appreciated that a harmful outcome was virtually certain. The AI engineer wants to classify a user’s communicative purpose from a string of text. These are not the same phenomenon wearing different hats. They are related but distinct aspects of a broader capacity: the ability to represent a goal, commit to a course of action, and organize behavior around that commitment.
What ties them together is the common thread of goal-directed commitment. Whether the timescale is milliseconds or months, whether the agent is an infant, a chimpanzee, a criminal defendant, or a software robot, intention marks the transition from passive awareness of possibilities to active engagement with one of them. It is the gear shift between thinking and doing, and it shows up in every domain where agents act in the world.