AUTOMAINTENANCE
Introduction to Automaintenance
The term automaintenance, within the field of behavioral psychology, refers specifically to the persistent continuation of a behavior that was originally established through a respondent or classical conditioning process, even when the maintenance of that behavior requires no contingency between the behavior and the subsequent reinforcing stimulus. This phenomenon is critical because it challenges traditional, rigid delineations between classical (respondent) and operant (instrumental) conditioning, illustrating a powerful interaction where behaviors are driven and sustained by stimulus-reinforcer pairings rather than response-outcome contingencies. Unlike purely operant behaviors, which necessitate that a response must occur to produce a consequence, automaintenance describes a situation where the behavior seems to be maintained automatically by the environmental predictability of the reinforcer, suggesting that the initial conditioned stimulus (CS) acquires potent eliciting properties that are highly resistant to standard extinction or punishment procedures.
The core example utilized to define and study automaintenance is the key-pecking behavior observed in pigeons, a response initially established through a procedure known as autoshaping, which is the spontaneous development of a conditioned response toward a stimulus that signals the impending delivery of a primary reinforcer, such as food. When this acquired behavior is placed under an omission schedule—a condition where the animal’s response actively prevents the delivery of the reinforcer—the bird often continues to perform the behavior, demonstrating the robust nature of automaintenance. This resistance to operant suppression highlights the dominance of the Pavlovian association in controlling the organism’s motor activity, compelling researchers to explore how non-contingent reinforcement can create behaviors that defy instrumental logic.
Understanding automaintenance is vital for developing comprehensive models of learning, particularly those that address species-specific defense reactions and biologically prepared behaviors, which often demonstrate a similar resistance to suppression or modification. The persistence of these behaviors suggests that internal motivational states and the predictive significance of environmental cues can override the immediate negative consequences associated with the behavior, forcing psychologists to consider a blended view of learning where classical and operant mechanisms frequently compete or cooperate to shape the final behavioral output. Thus, automaintenance serves as a crucial experimental preparation for isolating the powerful influence of stimulus-stimulus associations on behavior maintenance.
The Historical Context and Discovery
The phenomenon of automaintenance emerged directly from the seminal work on autoshaping conducted by Brown and Jenkins in 1968. Their original goal was simply to study the development of behavior under non-contingent procedures, where a key light (the CS) was presented for a brief duration and was invariably followed by the delivery of grain (the US), irrespective of what the pigeon was doing. What they observed was remarkable: the pigeons spontaneously began pecking the illuminated key. This discovery of autoshaping immediately presented a theoretical challenge, as the behavior developed without required shaping or differential reinforcement, suggesting a powerful role for classical conditioning in eliciting what appeared outwardly to be an operant response.
Following the initial acquisition phase (autoshaping), researchers naturally moved to examine the maintenance and persistence of this classically established response, leading to the identification of automaintenance. If the behavior was purely classical, it should extinguish relatively quickly if the stimulus-reinforcer pairing ceased or if the response was punished. However, when researchers introduced an omission schedule, which is the crucial test for automaintenance, the behavior proved remarkably durable. In these omission schedules, if the pigeon pecked the key, the food delivery that was scheduled to follow the key illumination was canceled or omitted. Logically, under instrumental learning principles, the behavior should rapidly decrease and extinguish because the response leads to punishment (the loss of the reinforcer), yet the pecking continued at high rates.
The historical significance of these findings lies in their implication that certain classes of learned behaviors are not solely determined by their consequences but are strongly dictated by the predictive power of environmental cues. Prior to this research, many learning theorists maintained a strict separation between reflexive classical responses and voluntary operant behaviors. Automaintenance demonstrated that a classically conditioned association (Key Light predicts Food) could generate and maintain an instrumental-looking response (Pecking), even when that response was detrimental to the organism’s immediate outcome, effectively blurring the line between the two primary forms of associative learning.
Mechanism of Autoshaping: The Precursor
To fully appreciate automaintenance, one must first understand the mechanism of autoshaping, which is the procedure that initially establishes the behavior. Autoshaping is fundamentally a Pavlovian process rooted in the concept of stimulus substitution or sign-tracking. When the conditioned stimulus (CS, the key light) is consistently paired with the unconditioned stimulus (US, the food), the CS begins to acquire some of the functional properties of the US. Because the food is an innate incentive, the light itself becomes an incentive or a ‘sign’ of the food, prompting the animal to interact with it in a manner appropriate to the US—in the pigeon’s case, pecking, which is a component of its species-specific feeding routine.
The conditioned response (CR) generated during autoshaping is therefore interpreted as a species-specific response system elicited by the sign-tracking stimulus. For instance, pigeons often peck the key with their mouths closed, resembling the ingestion phase of feeding, while they peck actual food with their mouths slightly open. This detail supports the idea that the response is not merely a generalized motor action but one specifically tied to the biological significance of the predicted outcome. The mechanism operates automatically because the associative learning occurs regardless of the animal’s behavior; the food delivery is time-locked to the light presentation, making the association entirely independent of the pecking response itself.
It is crucial to note that while autoshaping establishes the behavior, it does not necessarily explain the persistence of automaintenance under negative contingencies. The acquisition phase is about learning the predictive relationship (CS-US); the maintenance phase is about the durability of the resulting behavior when the environment actively attempts to suppress it via operant rules. The power of the initial classical association transforms the key light into a compelling eliciting stimulus, setting the stage for the behavior’s stubborn continuation during the omission trials that define automaintenance.
Defining Autoshaping vs. Automaintenance
While often discussed together, it is essential to draw a clear theoretical and procedural distinction between autoshaping and automaintenance. Autoshaping refers strictly to the acquisition phase: the process by which a behavior is spontaneously acquired due to the non-contingent pairing of a CS and a US. This phase is characterized by an increasing frequency of the conditioned response as the animal learns the predictive relationship between the stimulus and the reinforcer. The definition of autoshaping is rooted entirely in Pavlovian principles of association formation, where the animal learns that “X predicts Y.”
In contrast, automaintenance refers to the persistence of this behavior when an instrumental or operant contingency is introduced specifically to extinguish or suppress the response. The standard procedure for testing automaintenance involves the implementation of an omission schedule. Under this schedule, the CS (light) is still presented and followed by the US (food), but only if the animal refrains from responding. If the animal executes the conditioned response (e.g., pecking the key), the scheduled US is immediately canceled. This negative contingency creates a direct conflict: the animal is classically conditioned to respond to the sign, but operantly punished (by loss of food) for doing so.
The sustained high rate of responding observed during the omission schedule is the phenomenon of automaintenance. If the behavior were purely operant, it would rapidly extinguish because the response is reliably followed by the removal of the reward. The fact that the classically conditioned behavior persists, often at rates comparable to the acquisition phase, demonstrates the strength of the stimulus-reinforcer association and its ability to override the punitive effects of the response-outcome contingency. This distinction highlights that while autoshaping is the entry point, automaintenance is the critical evidence for the dominance of Pavlovian control over instrumental responding in this specific behavioral domain.
Theoretical Explanations and Models
The persistence of automaintenance has spurred significant theoretical debate, primarily revolving around whether the behavior is maintained by classical mechanisms (sign-tracking) or subtle, hard-to-control operant contingencies (adventitious reinforcement). The dominant theory supporting the classical view is the Stimulus-Substitution (SS) Theory, which posits that the CS (key light) acts as a literal substitute for the US (food). The animal interacts with the CS as if it were the US itself, and this interaction is inherently rewarding or biologically programmed. According to this perspective, the pecking behavior is not truly an instrumental action designed to produce food but rather an innate consummatory behavior elicited reflexively by the acquired properties of the CS. This model readily explains why the response resists operant suppression, as the response is driven by elicitation, not consequence.
A competing, though less favored, explanation involves the concept of superstitious reinforcement or adventitious reinforcement, which attempts to explain automaintenance through operant principles. This theory suggests that despite the omission schedule, there might be momentary lapses or specific temporal windows where an occasional response is inadvertently reinforced, perhaps due to slight timing errors or overlapping schedules, thus maintaining the overall response rate. However, meticulous experimental designs have largely refuted the sufficiency of this explanation, demonstrating that automaintenance persists even when reinforcement is systematically omitted following the response, making the classical explanation more compelling for the primary mechanism of maintenance.
Modern theoretical perspectives often favor an integrative approach, viewing automaintenance as a demonstration of the competition between classical and instrumental systems. According to this view, the Pavlovian system establishes a strong excitatory link (CS predicts US) that generates the response (pecking), while the instrumental system learns the negative contingency (Pecking omits US). The continued behavior indicates that the Pavlovian system is temporarily dominant or possesses a higher level of inertia, meaning the elicited response is so powerful that it requires a significant and prolonged instrumental suppression attempt to override the conditioned reflex. This dual-system perspective acknowledges the complexity of associative learning and the potential for biologically prepared responses to resist purely cognitive or instrumental control.
Experimental Evidence and Classic Studies
The primary body of experimental evidence supporting automaintenance stems from studies utilizing the pigeon key-peck preparation under omission schedules. In a typical experiment, pigeons are initially autoshaped until they reliably peck the key light when it signals food. Then, researchers introduce the omission contingency: if the bird pecks the key, the subsequent food is canceled. The key finding across numerous studies is the remarkable persistence of pecking behavior, often maintaining response rates exceeding 50% despite the fact that every peck results in the loss of a scheduled reinforcer.
Further critical evidence comes from comparisons involving different types of conditioned responses. For instance, studies have shown that if the conditioned response is topographically similar to the consummatory response (like the pigeon’s pecking resembling feeding), automaintenance is extremely robust. However, if the required response is arbitrary or distant from the consummatory behavior (e.g., if the animal had to step on a treadle to prevent the food omission), the arbitrary response is usually extinguished quickly under the omission schedule, behaving according to standard operant principles. This topographical specificity strongly supports the sign-tracking interpretation, indicating that automaintenance is specifically tied to responses that are biologically relevant to the predicted reinforcer.
Additionally, researchers have manipulated the parameters of the CS and US to investigate their impact on automaintenance rates. Factors such as the intensity of the CS, the temporal delay between CS and US, and the magnitude of the US all influence the strength of the underlying Pavlovian association and, consequently, the resistance of the resulting behavior to extinction via omission. These parametric studies consistently demonstrate that the variables which strengthen classical conditioning also enhance the degree of automaintenance observed, providing robust empirical support for the idea that the maintenance is fundamentally driven by the strength of the stimulus-stimulus association.
Implications for Learning Theory
The existence of automaintenance profoundly impacts contemporary learning theory by necessitating a modification of simplistic models that rigidly separate classical and operant conditioning. Historically, the two-factor theory maintained that classical conditioning governed involuntary, reflexive responses, while operant conditioning governed voluntary, skeletal responses. Automaintenance directly challenges this by showing that a seemingly voluntary, instrumental response can be sustained by purely classical, non-contingent relationships, even when it is counterproductive from an instrumental perspective. This forces theorists to acknowledge a significant domain where classical associations exert powerful control over instrumental actions.
Furthermore, automaintenance has been closely linked to the concept of the misbehavior of organisms, a term coined by Breland and Breland, describing how innate, species-specific behaviors can intrude upon and disrupt learned operant behaviors. Automaintenance provides a structured, laboratory-based example of this phenomenon: the classically elicited sign-tracking behavior (pecking) interferes with the operantly required behavior (withholding the peck to gain food). This highlights that learning is not a blank slate process but is constrained and guided by biological predispositions and the innate response systems specific to the species.
The theoretical implications extend to motivational psychology, suggesting that the conditioned stimulus, through its association with reward, acquires powerful incentive salience—the ability to motivate approach and interaction regardless of the current response-outcome relationship. The conditioned stimulus becomes a “wanted” object, and the response directed toward it is an instance of incentive motivation overriding rational instrumental behavior. Thus, automaintenance serves as a critical bridge, demonstrating how basic associative mechanisms lay the groundwork for complex motivational drives that sustain behavior even in the face of negative feedback.
Clinical and Applied Relevance
While initially studied in non-human subjects like pigeons, the principles underlying automaintenance possess significant clinical and applied relevance to human behavior, particularly in understanding the persistence of maladaptive or irrational habits that are resistant to logic or consequence. Many behaviors related to addiction, for example, can be viewed through the lens of automaintenance. Cues in the environment (CSs)—such as seeing drug paraphernalia, entering a specific location, or interacting with certain people—become powerfully associated with the drug effect (US). These cues acquire incentive salience and elicit a craving or approach behavior (the CR).
In the context of addiction, the approach or consummatory behavior might be maintained even when the individual is consciously aware that the behavior leads to severe negative outcomes (the omission schedule equivalent). The classical association (Cue predicts Reward) is so strong that it overrides the instrumental knowledge (Response leads to Punishment), resulting in the persistent pursuit of the substance. Similarly, certain compulsive or ritualistic behaviors, phobias, and even eating disorders might involve elements of automaintenance, where the behavior is maintained by the eliciting properties of specific environmental or internal cues rather than the immediate positive consequences of the action.
The application of automaintenance principles suggests that effective therapeutic interventions must go beyond simply punishing the maladaptive response or highlighting the negative consequences (pure operant approaches). Instead, effective strategies should focus on extinguishing the underlying Pavlovian association by repeatedly presenting the conditioned stimulus without the unconditioned stimulus, thereby reducing the incentive salience of the cue itself. For example, exposure therapy aims to decouple the fearful cue from the traumatic outcome, directly attacking the classical association that drives the persistent avoidance behavior.
