DRAGENwiki · Article

Determinism

Sources checked. Clinical review pending.

Determinism is the idea that events, behavior included, happen because of what came before. Behavior analysis works from that assumption: behavior has reasons we can look for in a person's biology, past learning and present situation. That does not make behavior easy to predict, and it does not mean people have no choices. Behavior analysts themselves still argue about how far it goes.

Determinism in 60 seconds

  • The idea. Roughly speaking, determinism says every event is "necessitated by antecedent events and conditions together with the laws of nature." [1]
  • Behavior analysis assumes it to do its work. In 1953, B. F. Skinner wrote that to use the methods of science in human affairs, "we must assume behavior is lawful and determined." [2]
  • It is not fate. Fatalism says some events are destined to occur "no matter what we do." Determinism is about events unfolding through natural laws and cause and effect, and in a deterministic world nothing has to be fated. [1]
  • It does not promise exact prediction. Even in the lab, some behavior varies so much that only its averages can be predicted. [2]
  • It sits next to choice. A 2025 review argues that although behavior is determined, people "have opportunities to make choices and exercise autonomy." [4]
  • It is a working tool, not a verdict on a person. At home and school, it turns "Why would anyone do that?" into "What made this more likely, and what can we change?" A BCBA reviews every clinical decision.

Jump to: What determinism means · Why behavior analysts assume it · Probability, not clockwork · A debate in print · Free will · At home and school · A made-up example · Myths · Words to know · Sources

General information, not medical or clinical advice.

What determinism means

The Stanford Encyclopedia of Philosophy describes causal determinism as the idea that every event follows from earlier events and conditions plus the laws of nature. The idea is ancient, but thinkers first began to clarify it and analyze it mathematically in the eighteenth century. [1]

A famous picture of it comes from Laplace. In 1820 Laplace imagined an intelligence that knew every force acting in nature and the position of every thing at one moment. To such a mind, Laplace wrote, nothing would be uncertain: the future and the past would both lie open. The encyclopedia retells the story with "a sufficiently bright demon" in the role of that mind. [1]

The encyclopedia separates determinism from two ideas it is often confused with. [1]

  • Fate. Fatalism is the claim that some or all events are destined to happen "no matter what we do." Its guarantee comes from a divine will or foreknowledge, or from a built-in purpose in the universe, not from events unfolding under natural laws and cause and effect. The encyclopedia notes that we can imagine a world run by deterministic laws in which nothing at all is fated to occur. [1]
  • Prediction. Studies of chaos show that even predicting one small, well-defined part of the world for a limited time may be impossible for a finite agent living in that world. For humans and other finite agents, the encyclopedia concludes, "predictability and determinism are simply not logically connected at all." [1]

In physics and in thinking about free action, there is still no agreement over whether determinism is true, or even whether it could be known to be true or false. [1] Behavior analysis does not need to settle that question to do its work, as the next sections show.

Why behavior analysts assume it

Behavior analysis is a natural science of behavior. To study anything scientifically, you have to expect that it follows regularities you can discover. Skinner said this plainly in Science and Human Behavior (1953): to use the methods of science in human affairs, "we must assume behavior is lawful and determined." [2]

In practice, the assumption is a promise to keep looking. When a behavior happens, a behavior analyst asks what in the person's history and present situation made it more likely. A 2025 review by Richard Rakos, Ruth Anne Rehfeldt and Peter Killeen puts it simply: "The whole purpose of functional analysis is to find those factors." [4] Results are never perfect, the authors add, but causes can often be engaged to change behavior. They write that they are scientists and practitioners because they can step in and change things. [4]

The same review traces this attitude to the Enlightenment, when a young science needed nature to be orderly enough to predict. The authors sum it up in two words, "Assume orderliness!", and call the stance pragmatic. [4]

Not everyone agrees that science needs the assumption at all. Marr quotes Day (1969), who wrote that radical behaviorists look for functional relationships without assuming in advance a universe that is "orderly in a mechanical sense." [2] And William Vaughan reported that in a 1951 letter to Skinner, Percy Bridgman said science need not assume the universe is lawful. It need only use the lawful relations it has already found. [3] On that view, determinism is less a belief about the universe and more a habit: keep looking for the conditions that matter, and use the ones you find.

Probability, not clockwork

It would be easy to picture determinism as a clock, where every gear moves the next one exactly. Jack Marr, writing in The Behavior Analyst in 1982, argued that a closer look at the field's principles and practice shows a more complex picture, in some ways closer to the physics of Bohr and Heisenberg than to that of Newton and Laplace. [2]

  • Probability is central. The probability of a response is a central idea in behavior analysis. Marr noted it is unclear whether that reflects only our inability to specify every starting condition, or partly random processes in how behavior is emitted. [2]
  • Operant behavior is "emitted." Skinner contrasted operant behavior with reflexes, which are elicited by a stimulus. For an operant, Skinner wrote in 1974, "there seems to be no necessary prior causal event." [2]
  • Some variation looks random, even in the lab. Under fixed-interval schedules, pause times and response counts appear to be randomly distributed. Marr wrote that "one is only able to predict the average values of output." [2]
  • Studying a process can change it. To learn what controls performance on a schedule, the usual tactic is to change the schedule. But turning an indirect variable into a direct one will likely change performance without necessarily showing how, or even if, that variable works indirectly. [2]

None of this means anything goes. Physicists cannot change how fast a radioactive substance decays, Marr noted, but conditioning is defined by the fact that certain arrangements can change the probability of behavior. Unlike atoms, operants have a "sense of history": what happened before changes what happens next. [2] Membership in a species and each individual's genes also set natural limits on the form and variability of emitted behavior. [2]

Marr also rejected the idea that this kind of spontaneity rescues "free will." A behavior that simply appears is no more under the control of the one who emits it than it is under the control of an observer. Spontaneity, Marr wrote, "can provide no solace" to defenders of free will. [2] Marr's conclusion was hopeful: giving up mechanistic, clockwork determinism "should not be viewed by behaviorists with despair, but rather be looked upon as liberating." [2]

A debate in print

Marr's article drew a reply the next year. William Vaughan, Jr., then at Harvard University, asked what the evidence for spontaneity really was. Vaughan's answer: "It is, and could only be, lack of evidence." [3] When a behavior changes and nobody can name its controlling variables, that shows only what we have not found yet.

Vaughan put it this way: "if controlling variables are found, then they exist." If they are not found, no answer is possible unless someone has searched every relevant variable. [3] Would more knowledge of the organism's history help? More knowledge of the current environment? Recordings from single nerve cells, or more powerful computers? Calling a behavior spontaneous, Vaughan argued, amounts to inferring that none of these would improve prediction. [3]

Vaughan offered two guesses about why a scientist might say "spontaneity." The first is the same thing Skinner said keeps everyday explanations going: once the word is said, "further inquiry is halted." [3] The second is reluctance to accept Laplace's idea that knowing the whole universe at one moment would reveal every past and future state, our own actions included. If spontaneity is real, the truth lies somewhere between free will and hard determinism, and since hard determinism provokes revulsion while free will seems absurd, a compromise can look sensible. [3]

The exchange matters for families and teachers. When a behavior seems to come "out of nowhere," Vaughan's point is to keep looking. Marr's point is to stay humble about how much any one moment can be predicted. Both lead to the same daily habit: take good data over many days, and look for patterns instead of explaining a single moment.

Determinism and free will

The free will debate is much older than behavior analysis. Rakos, Rehfeldt and Killeen (2025) describe three positions that have come to dominate it, plus a newer fourth. [4]

  • Libertarianism (about free will, not politics) says free will is real and is not compatible with determinism. [4]
  • Hard determinism says that because determinism is true, people lack the kind of free will that moral responsibility would require. The review names the neuroscientist Robert Sapolsky (2023) and Skinner's books of 1971 and 1974 as holding this view. [4]
  • Compatibilism, sometimes called soft determinism, says free will can exist in a deterministic universe. [4]
  • Revisionism, the newer fourth view, also says free will can exist in a deterministic universe. It treats free will as a kind of social skill: recognizing and responding well to what matters in a situation. The review calls this view friendly to behavior analysis. [4]

These positions lead to different views about blame. Libertarians, compatibilists and revisionists keep blame and credit for actions. According to the review, only hard determinists reject the kind of moral responsibility that justifies punishment. They prefer "contingent accountability that promotes rehabilitation," which aims at change rather than punishment. [4]

Behavior analysts do not all agree, either. When a recent behavior-analytic symposium polled its audience, the review reports, the overwhelming majority voted that behavior is determined; only one person of perhaps 100 voted "free." [4] The three authors themselves say they take differing positions. [4] The view they offer is called probabilistic determinism: the past is determined, but the future can only be predicted in broad terms, because people weigh many considerations as they choose among the options open to them. Behavior is determined, they suggest, yet people still make choices and exercise autonomy. [4]

People think about these questions inside many faiths and philosophies, and outside any. Behavior analysis uses determinism as a working method for science. It does not ask anyone to give up what they believe about God, as each person understands God, or about meaning and purpose.

What it means at home and school

Determinism becomes useful when it changes what adults do next.

  • Swap blame for curiosity. "Lazy," "stubborn" and "doing it on purpose" describe a person, not a reason. Ask instead: what happened right before, what happened right after, and what might make this behavior more likely today? DRAGEN's lessons on the four-term contingency and the four functions of behavior show how to look.
  • Change the situation, not the person's worth. If behavior has causes, supports work by changing conditions: the task, the setting, the help available and what follows. Finding those conditions is what a functional analysis is for, and a BCBA leads that work. [4]
  • Teach more options. Rakos and colleagues write that their view implies prosocial actions are more likely when people have a wide range of capabilities for responding flexibly and effectively to surprising and uncertain situations. [4] Teaching a child to ask for help, a break or a different task gives them more ways to meet their needs.
  • Offer real choices. In the 2025 review's view, determinism and choice can live together. [4] Choices about order, place, materials and helpers respect a person's autonomy, and a person's choices are part of what shapes what they do next.
  • Treat "no" as information. A refusal has reasons too. It might say a task is too hard, too long, painful or not meaningful to the person. DRAGEN treats "no," refusal and withdrawn assent as information to listen to, never as a target to reduce. When there is an immediate safety need, adults act to keep everyone safe, and that is handled separately from a person's dissent.
  • Look at many days, not one moment. Because single moments can be hard to predict, patterns over time are more trustworthy. [2] A graph of the behavior across days shows what one hard afternoon cannot.
  • Hold accountability without punishment. Accountability can focus on rehabilitation, as hard determinists propose. [4] In daily life, that can mean repairing, practicing and trying again with better support. DRAGEN plans focus on teaching and support, and a BCBA reviews every clinical decision.

A made-up example

This example is invented. It does not describe a real person.

Learner F is 10. On most school nights, when homework comes out, F says "No" and goes to the couch. A tired parent thinks, "F is choosing to be difficult."

With a BCBA's help, the family tries the determinism habit: assume there are reasons, and look for them.

  1. Write it down. For two weeks, the family notes what happened before homework, what the homework was and what happened after F said no. They use a simple four-term data sheet.
  2. Look for patterns. F starts within a few minutes on nights with one short page. F says no on nights with long writing pages, especially after sports practice.
  3. Ask F. F says writing makes their hand hurt and long pages feel like they never end. The family treats this as information, not an excuse.
  4. Change the conditions. The family splits long pages into short parts with a check mark after each one. F picks which part to do first and where to sit. A break card is always on the table, and asking for a break always works.
  5. Teach a new option. The family practices saying "I need help" or "Break, please" at calm times.
  6. Keep counting. After a month, F starts homework on more nights. When F still says no, the family asks what is hard and adjusts. The family also tells the teacher about the writing pain, and the teacher and family decide together whether to involve a doctor or therapist.

Nobody decided that F had no will of their own. The family decided the reasons could be found and changed. A BCBA reviews each step.

Common myths

  • Myth: Determinism means everything is fate. Fact: Fatalism says events will happen no matter what anyone does. Determinism is about events unfolding through natural laws and cause and effect, and a deterministic world need not have anything fated at all. [1]
  • Myth: If behavior is determined, nobody can choose. Fact: Several positions in the debate hold that free will can exist in a determined world. A 2025 review argues that people still make choices and exercise autonomy. [4]
  • Myth: Behavior analysts think people are machines. Fact: Marr argued that the field's picture is closer to probability than to clockwork. Some behavior can only be predicted on average. [2]
  • Myth: Determinism means we can predict exactly what someone will do. Fact: Even predicting a small part of the world for a limited time may be impossible for finite agents like us. For us, predictability and determinism are separate ideas. [1]
  • Myth: If behavior has causes, nobody is responsible. Fact: Three of the four main positions keep blame and credit for actions. Even hard determinists keep accountability, in the form of "contingent accountability that promotes rehabilitation." [4]
  • Myth: All behavior analysts agree about determinism. Fact: Marr and Vaughan disagreed in print, and in 2025 three authors said they hold differing positions. [2][3][4]

Words to know

  • Determinism: the idea that every event follows from earlier events and conditions together with the laws of nature. [1]
  • Fatalism: the idea that some or all events are destined to happen no matter what anyone does. [1]
  • Laplace's demon: the imagined mind in Laplace's story, which knows the present state of everything and could calculate the whole past and future. [1]
  • Hard determinism: the view that every event, behavior included, follows inevitably from its causes, so people lack the kind of free will that moral responsibility would require. [4]
  • Compatibilism (soft determinism): the view that free will can exist in a deterministic universe. [4]
  • Libertarianism (free will): the view that free will is real and is not compatible with determinism. [4]
  • Revisionism: a newer view that treats free will as a social skill: recognizing and responding well to what matters in a situation. [4]
  • Probabilistic determinism: the view that the past is determined while the future can only be predicted broadly, because people choose among options. [4]
  • Emitted behavior: operant behavior, as opposed to a reflex that is elicited by a stimulus. [2]
  • Spontaneity: behavior that seems to occur without an identifiable cause; behavior analysts disagree about whether it is real. [2][3]
  • Controlling variable: a condition that a behavior depends on. [3]
  • Functional analysis: the work of finding the causal factors that a behavior depends on. [4]

Need a PDF?

Create it here, for free: the DRAGEN app. Use it at your next session.

Or print one we already made, also free:

More from DRAGEN

Related DRAGENwiki pages: behaviorism · radical behaviorism · methodological behaviorism · experimental analysis of behavior · seven dimensions of ABA · applied behavior analysis · four-term contingency · reinforcement · motivating operations · the right to say no · reading a behavior graph

What remains unknown

  • Whether determinism is true is unsettled, in physics and in philosophy, and so is whether it could ever be known. [1]
  • Whether behavior that looks spontaneous is truly random, or only reflects variables nobody has found yet, is still argued. Marr and Vaughan took opposite sides in 1982 and 1983. [2][3]
  • The free will debate itself is unresolved. The 2025 review's authors note how long it has lasted and that there seems to be no way to resolve it to everyone's satisfaction, and they take differing positions themselves. [4]
  • This page summarizes the main lines. It does not settle these debates.
  • Last checked on November 12, 2026.

Sources

  1. Hoefer, C. (2023). Causal determinism. In The Stanford Encyclopedia of Philosophy (substantive revision of September 21, 2023; first published January 23, 2003). https://plato.stanford.edu/entries/determinism-causal/
  2. Marr, J. (1982). Determinism. The Behavior Analyst, 5(2), 205–207. https://doi.org/10.1007/BF03392389 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC2742056/)
  3. Vaughan, W., Jr. (1983). Comments on Marr's determinism. The Behavior Analyst, 6(1), 111. https://doi.org/10.1007/BF03391880 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC2742073/)
  4. Rakos, R. F., Rehfeldt, R. A., & Killeen, P. R. (2025). The free will: Determinism debate in contemporary science and society (or: Why are you reading this article?) [Review of the books Determined: A science of life without free will, by R. Sapolsky, and Free agents: How evolution gave us free will, by K. Mitchell]. Perspectives on Behavior Science, 48(2), 475–497. https://doi.org/10.1007/s40614-025-00447-3 (full text: https://pmc.ncbi.nlm.nih.gov/articles/PMC12162392/)

Pages that link here