The big idea
Every science starts from assumptions it cannot prove but needs in order to work. Behavior analysis assumes that behavior is lawful, that we learn about it by observing it, and that simple, testable explanations should be ruled out before complex ones. These assumptions shape how a behavior analyst asks questions, collects data and judges an explanation.
The outline names five: selectionism, determinism, empiricism, parsimony and pragmatism. Cooper, Heron and Heward (2020) list a similar set as the attitudes of science: determinism, empiricism, experimentation, replication, parsimony and philosophic doubt. Know the outline’s five well and the others by name, because questions can draw on either list.
None of this is abstract in practice. These assumptions decide whether you blame a child’s “bad attitude” or look at what happens after the behavior, and whether you trust a testimonial or a graph.
Breaking it down
Selectionism
Selectionism holds that behavior, like a species, is shaped by its consequences. Variation occurs, and some variations are selected because of what they produce. Skinner (1981) described selection working at three levels:
- Phylogenic: natural selection across generations shapes the species, including reflexes and the capacity to be changed by consequences.
- Ontogenic: within one lifetime, consequences select some responses over others. Operant learning is selection at this level.
- Cultural: practices passed among members of a group persist or fade depending on their effects on the group, such as a hospital’s handwashing routine.
Selectionism is why behavior analysts look after a behavior, as well as before it, for its causes.
Determinism
Determinism is the assumption that the universe, behavior included, is lawful and orderly. Events have causes, and behavior is a function of genetic endowment, learning history and the current environment. Behavior analysts do not prove determinism; they assume it, because a science of random events would be impossible.
Determinism is not fatalism. If behavior has causes, it can be changed by changing those causes, which is the optimistic heart of applied work.
Empiricism
Empiricism means knowledge comes from objective observation and measurement, not from opinion, authority, tradition or reasoning alone. A behavior analyst trusts what is measured over what people say happened. When a parent reports that a new diet ended their child’s tantrums, the empirical response is to measure the tantrums.
Parsimony
Parsimony says to rule out the simplest, most testable explanations before turning to more complex or abstract ones. If a student’s refusals can be explained by escape from hard work, there is no need yet to invoke an unobservable inner state. Parsimony is not a claim that behavior is simple. It is a rule about the order in which explanations are tried.
Pragmatism
Pragmatism judges an explanation by how well it works: does it help predict and change behavior? An account is useful, and in that sense true, when acting on it produces the expected result. A functional assessment earns its value when the intervention built on it succeeds.
The other attitudes of science
- Experimentation: manipulating variables to find out what causes what. It is the basic strategy for reaching control (see A.1).
- Replication: repeating an experiment, or a condition within one, to see whether the result holds. Findings that replicate earn more confidence.
- Philosophic doubt: holding every finding, including your own, as tentative and open to revision when new data arrive.
Seeing it in practice
A middle school asks a behavior analyst to help with Diego, who leaves class several times a day.
| Assumption | How it shows up |
|---|---|
| Determinism | The analyst assumes Diego’s leaving has identifiable causes and is not random. |
| Empiricism | Instead of relying on staff impressions, the analyst records each exit and what happened just before and after it. |
| Selectionism | Leaving is usually followed by a walk with a favorite aide, so the analyst asks what that consequence has selected. |
| Parsimony | Before appealing to an inner trait such as defiance, the analyst tests the simpler idea that leaving is maintained by escape and attention. |
| Pragmatism | The explanation is judged by results. A plan built on it reduces exits, which supports the account. |
| Philosophic doubt | When exits creep back up in spring, the analyst reopens the question instead of assuming the old answer still holds. |
Common mix-ups
- Determinism is not fatalism. It does not mean a person is doomed by their history. It means behavior has causes that can be identified and changed.
- Parsimony is not “the simplest answer is always true.” It sets the order for testing explanations. A complex explanation is fine once the simpler ones have been ruled out.
- Empiricism is not the same as experimentation. Empiricism bases claims on observation. Experimentation adds the manipulation of variables. A careful descriptive study is empirical without being an experiment.
- Pragmatism and philosophic doubt are different questions. Pragmatism asks whether an explanation works. Philosophic doubt asks whether you should keep believing it.
- Assumptions are not goals. Description, prediction and control (A.1) are what the science aims to achieve. These assumptions are what it takes for granted along the way.
Exam tip: Match the key phrase to the assumption. Lawful, orderly, caused signals determinism. Observed, measured, objective signals empiricism. Simplest testable explanation first signals parsimony. Judged by whether it works signals pragmatism. Consequences select signals selectionism. Tentative, open to revision signals philosophic doubt.
Quick recap
- Selectionism: consequences select behavior at three levels (phylogenic, ontogenic, cultural).
- Determinism: behavior is lawful and caused. Empiricism: knowledge rests on objective observation.
- Parsimony: rule out simple, testable explanations first. Pragmatism: judge an explanation by how well it works.
- Cooper et al. also list experimentation, replication and philosophic doubt as attitudes of science.
Further reading
- Cooper, J. O., Heron, T. E., & Heward, W. L. (2020). Applied behavior analysis (3rd ed.), chapter 1.
- Skinner, B. F. (1981). Selection by consequences. Science, 213(4507), 501-504.
- Baum, W. M. (2017). Understanding behaviorism: Behavior, culture, and evolution (3rd ed.). Wiley-Blackwell.
Key terms
- Selectionism
- The assumption that behavior, like the features of a species, is shaped by its consequences: variations occur, and those that produce certain outcomes are selected. Skinner described selection at three levels: phylogenic (natural selection), ontogenic (operant learning within a lifetime) and cultural (practices passed among members of a group).
- Determinism
- The assumption that the universe, behavior included, is lawful and orderly, so that events occur as a result of other events. Behavior is treated as a function of genetic endowment, learning history and the current environment, which means it can be changed by changing those variables.
- Empiricism
- The attitude that knowledge should rest on objective observation and measurement of events, rather than on opinion, authority, tradition or reasoning alone.
- Parsimony
- The practice of ruling out the simplest, most testable explanations for behavior before considering more complex or abstract ones. It sets the order in which explanations are tested; it is not a claim that behavior is simple.
- Pragmatism
- The view that an explanation is judged by how well it works: whether acting on it leads to successful prediction and change of behavior.
- Philosophic doubt
- The attitude of treating all findings, including one's own, as tentative and open to revision in light of new data. It keeps a scientist asking whether what is believed to be true actually is.
- Experimentation
- The basic strategy of science: deliberately manipulating one variable while measuring another, to find out whether changes in the first produce changes in the second. It is how a functional relation is demonstrated.
- Replication
- Repeating an experiment, or a condition within an experiment, to see whether the same result occurs again. Within a study it strengthens the case for a functional relation; across studies it shows how reliable and general a finding is.
Check yourself
5 questions on A.2. Choose an answer to see whether you're right and why.
Show the answer
Answer: A. Determinism is the assumption that behavior is lawful and has causes. The analyst has not measured anything yet, so the point is not empiricism, which concerns how knowledge is gained. Parsimony would come in later, when choosing which possible cause to test first.
Show the answer
Answer: B. Parsimony means ruling out the simplest, most directly testable explanation before turning to complex or inferred ones. Philosophic doubt is about staying open to revising accepted findings, not about the order in which explanations are tested.
Show the answer
Answer: D. Pragmatism judges an explanation by how useful it is for predicting and changing behavior, and the analyst's argument rests on what the account let the team do. Empiricism is involved, since the result was measured, but it is not the point the analyst is making.
Show the answer
Answer: A. Ontogenic selection is operant selection within one lifetime: the parents' reactions selected the toddler's babbling. The reflex and the beak length are products of phylogenic selection across generations, and the checklist is a cultural practice kept because of its effects on the group.
Show the answer
Answer: B. Philosophic doubt means holding findings, even well-supported ones, as open to revision when new data arrive. Keeping the system because of past success treats an earlier result as settled truth, which is the opposite of this attitude.
Adds a step to your study record. Saved in this browser only.