In this foundational 1935 paper, B.F. Skinner—one of the most influential figures in 20th-century psychology—lays the groundwork for modern behavioral science. He argues that the term "conditioned reflex" is being used too loosely and proposes a rigorous formal distinction between two fundamentally different learning processes: what we now recognize as Operant Conditioning (Type I) and Classical/Pavlovian Conditioning (Type II).
TL;DR
Skinner formalizes the distinction between behavior that produces a consequence (Type I) and behavior that is elicited by a preceding stimulus (Type II). He argues these are distinct biological mechanisms with different rules for reinforcement, drive association, and measurement.
What Problem Does This Paper Address?
In the early 1930s, the study of "conditioning" was dominated by Pavlov’s work on salivating dogs. However, researchers were struggling to fit all learned behaviors into Pavlov’s stimulus-response (S-R) model. Some behaviors seemed to be "voluntary" or "purposive," aimed at achieving a goal, while others were "involuntary" physiological reactions.
Skinner addresses this conceptual muddle by defining two distinct paradigms. He seeks to prove that "Type I" (where a response is required to get a reward) is not just a variation of "Type II" (where a reward follows a stimulus regardless of behavior), but a separate category of reflex altogether.
Methods
This is primarily a theoretical and taxonomic paper. Skinner uses symbolic logic to map out the temporal and causal relationships between stimuli ($S$) and responses ($R$).
He defines the paradigms as follows:
- Type I (Operant): $S_0 \rightarrow R_0 \rightarrow S_1$. Here, the response ($R_0$) must occur for the reinforcing stimulus ($S_1$) to appear. The goal of conditioning is to increase the strength of the $[S_0 - R_0]$ bond.
- Type II (Respondent): $S'_0 \rightarrow S'_1$. Here, the two stimuli are paired regardless of the organism's behavior. The goal is for the first stimulus ($S'_0$) to eventually elicit the response ($R'_1$) normally associated with the second stimulus.
Datasets
As a theoretical paper, it does not present new experimental data. Instead, it synthesizes and re-interprets established laboratory observations from the era, such as rats pressing levers for food (Skinner's own work) and dogs salivating to bells (Pavlov's work).
Key Insights and Ideas
- The Necessity of the Response: In Type I, the response is the "bridge" to reinforcement. If the rat doesn't press the lever, it doesn't get food. In Type II, the response is a byproduct; the dog gets food after the light whether it salivates or not.
- Drive Flexibility vs. Fixity: Skinner notes a fascinating difference in "drives" (like hunger or thirst). Type I is drive-agnostic: you can train a lever press to be fueled by hunger or thirst. Type II is drive-locked: a light paired with food will only ever elicit salivation when the animal is hungry.
- The "Pseudo-Type" Warning: Skinner identifies a "pseudo-conditioned reflex" which is actually a discrimination task. For example, if a rat only presses a lever when a light is on, the light isn't a Pavlovian "conditioned stimulus" for pressing; rather, the rat has learned that the $[S_{lever} - R_{press}]$ reflex is only reinforced in the presence of the light.
- Biological Economy: Skinner assigns different roles to each type. Type II is for preparation (e.g., salivating to prepare for digestion). Type I is for acquisition (e.g., the actual behavior required to find and seize food).
Key Results
- Formal Taxonomy: Skinner successfully differentiated the two types across seven dimensions, including the origin of the reflex (Type I starts from "investigatory" behavior, Type II can start from zero) and the nature of the response (Type II is limited to simple glandular or local muscular actions).
- Measurement Differences: He points out that Type I strength can be measured continuously, whereas measuring Type II requires withholding the reinforcement, which inherently causes "extinction" (the weakening of the reflex).
- Defining the "Operant": While he uses the term "Type I" here, this paper is the birth of the "Operant" concept—behavior that "operates" on the environment to produce effects.
Limitations and Future Work
Skinner acknowledges that in the real world, these two types are often entangled. For instance, when a dog is trained to salivate to a light (Type II), it might also learn to turn its head toward the light (Type I) if that orientation is somehow reinforced.
He also notes that Type II is difficult to observe in a "pure" state in the laboratory because background stimuli (the "experimental stand") often become part of a complex discriminative environment. This paper set the stage for Skinner's later, more exhaustive work in The Behavior of Organisms (1938).
Summary updated