Learned Helplessness in Dogs: Loss of Control, Behavioral Suppression, and Pathways to Recovery

1. Introduction and Historical Background

1.1 Seligman and Maier (1967): The Original Experiments

The discovery of learned helplessness emerged from research that was not originally designed to study it. Seligman and Maier (1967) were investigating classical conditioning and avoidance learning when they observed an unexpected pattern: dogs that had previously been exposed to inescapable electric shock, delivered while restrained in a harness, subsequently failed to learn to escape shock in a shuttle box – a different apparatus in which escape was both possible and easily learned by naive dogs.

The experimental design involved three groups. One group of dogs could escape shock in the harness by pressing a panel with their nose. A second, "yoked" group received shocks of identical timing and duration but had no control over shock termination – their shocks were terminated only when the first group's dogs pressed the panel. A third group received no shock at all. When subsequently tested in a shuttle box, where shock could be escaped or avoided by jumping over a barrier, dogs from the escapable-shock and no-shock groups readily learned to escape and avoid. Dogs from the inescapable-shock (yoked) group, by contrast, largely failed to escape: many simply lay down and passively endured the shock, even when escape required only a simple, easily learned response (Seligman & Maier, 1967).

The critical theoretical claim was that the relevant variable was not the shock itself – both the escapable and inescapable groups received physically identical shock – but the controllability of the shock. The yoked dogs had learned, according to the original interpretation, that their behavior was independent of the outcome: nothing they did mattered. This learned expectation of response-outcome independence was proposed to generalize to the new situation, undermining the motivation to attempt escape even when escape had become possible.

Subsequent work by Seligman, Maier, and Geer (1968) demonstrated that this learned passivity could be reversed – "cured" – primarily through forced exposure to successful escape, in which experimenters physically guided dogs across the shuttle box barrier until the dogs began to make the response voluntarily. This early therapeutic finding foreshadowed a theme that remains central to contemporary discussions of recovery: passive exposure to the absence of shock was insufficient for recovery; what mattered was the dog's own experience of producing an escape response that worked.

It is important to note for the purposes of this review that these foundational experiments involved procedures – the deliberate exposure of dogs to inescapable electric shock – that would not meet contemporary ethical standards for animal research and that are entirely distinct from, and should not be confused with, the welfare-oriented applications of learned helplessness theory discussed in this article. The theory's origins in this research are part of its history and are relevant to understanding its development, but they are not endorsed or replicated in any context relevant to modern dog training or behavior therapy.

1.2 Definition and Initial Theoretical Claims

In its original 1967 formulation, learned helplessness was defined as a state of passivity and reduced responding resulting from prior experience with uncontrollable aversive events – outcomes that occur independently of the organism's behavior. The theory proposed three core deficits that result from this learning: a motivational deficit (reduced initiation of voluntary responses), a cognitive deficit (impaired learning of new response-outcome contingencies, even when they exist), and an emotional deficit (depression-like affective changes).

The theory was extended to humans by Seligman and colleagues throughout the 1970s, most influentially in the "attributional reformulation" by Abramson, Seligman, and Teasdale (1978), which proposed that the way an individual explains the causes of uncontrollable events – their attributional style – determines the breadth, duration, and severity of helplessness deficits. This reformulation became a cornerstone of cognitive theories of depression in humans and remains influential in clinical psychology, though it is largely outside the scope of this article's focus on dogs.

1.3 The 2016 Reformulation: "Learned Helplessness at Fifty"

Fifty years after the original experiments, Maier and Seligman (2016) published a substantial revision of the theory based on intervening decades of neuroscience research – much of it conducted by Maier's own laboratory using rodent models. The central claim of this reformulation is striking: the original theory, in a specific and important sense, had the causal direction backward.

According to the revised account, passivity in response to prolonged, intense, or inescapable aversive events is not learned. It is the default, unlearned response of the mammalian nervous system to such events, mediated by serotonergic neurons in the dorsal raphe nucleus (DRN), whose heightened activity during inescapable stress promotes behavioral passivity and what the authors describe as an anxiety- and fear-like state. What requires learning, in this revised account, is not passivity but its opposite: the detection of control. When an organism learns that its responses do affect outcomes, this is detected by circuitry in the medial prefrontal cortex (mPFC), which then actively inhibits the dorsal raphe nucleus, blocking the default passivity response and permitting active, exploratory, and coping behavior.

This reformulation has several important implications. First, it reframes "helplessness" not as a pathological state that must be unlearned, but as the default state from which active coping must be learned – or, in cases of prior successful coping, from which the expectation of control must be retained and generalized. Second, it situates learned helplessness within a broader neurobiological framework of stress, fear, and anxiety circuitry rather than treating it as a separate, isolated phenomenon. Third, it provides a mechanistic account of why experiences of control – even control over events unrelated to a current stressor – can have protective, "immunizing" effects on subsequent stress responses, a phenomenon termed behavioral immunization in the rodent literature (Amat et al., 2010).

For dogs, the 2016 reformulation has not been directly tested, but its implications are substantial. If passivity is the default response to prolonged uncontrollable stress, then the appropriate framing for behavior modification is not "removing learned passivity" but "building and generalizing the neural machinery of control detection" – a framing that aligns closely with contemporary emphases on agency, choice, and predictability in welfare-oriented dog training, discussed in Section 7.

For background on the neurochemical systems involved in learning and stress in dogs, see Dopamine and Learning in Canine Neurochemistry and Hormones in Dogs: How Neurochemistry Shapes Behavior.

Why This Matters for Dogs Today

Although the original experiments would be considered unethical by contemporary standards and are not part of modern research practice, the central question they raised remains directly relevant to canine welfare today: how do experiences of uncontrollability and unpredictability – whether from aversive training methods, chronically restrictive housing, or simply unstructured and inconsistent daily environments – influence a dog's learning, behavior, and emotional wellbeing? A dog that has stopped trying to escape a stressful situation, that no longer attempts to solve simple problems, or that appears to have "given up" on a behavior it once performed reliably may be displaying a pattern that this body of theory was specifically developed to explain. Understanding the mechanisms reviewed in this article – and their limits – is therefore not an academic exercise but a foundation for recognizing, preventing, and addressing one of the more serious and frequently overlooked welfare concerns in companion dogs.

2. Theoretical Framework

2.1 Uncontrollability Theory

The central theoretical variable in learned helplessness research, both in its original and revised forms, is controllability – the degree to which an organism's behavior influences the occurrence, timing, intensity, or termination of events in its environment. Critically, controllability is distinct from the aversiveness of the event itself. The classic yoked-control design isolates this variable precisely: two groups of animals receive physically identical aversive stimulation, differing only in whether their behavior affects that stimulation. The consistent finding, replicated across rodents, dogs, and other species, is that the yoked (uncontrollable) group shows substantially greater behavioral, physiological, and neurochemical disruption than the group with control, despite identical physical stimulation (Maier & Seligman, 2016).

This finding has a direct corollary for applied contexts: the welfare impact of an aversive event cannot be assessed from its physical properties alone. An identical stimulus – a sound, a physical sensation, a restriction of movement – may have dramatically different effects depending on whether the individual experiencing it has any behavioral influence over its occurrence or termination.

2.2 Unpredictability and Its Relationship to Uncontrollability

Closely related to, but conceptually distinct from, controllability is predictability – whether an organism can anticipate the occurrence of an event based on available cues. Uncontrollable events are often also unpredictable, and disentangling the two variables has been a persistent challenge in the experimental literature. However, the available evidence suggests that both variables contribute independently to stress responses, and that unpredictability may carry additional welfare costs even when some degree of control is present.

This is directly relevant to dogs. Schalke et al. (2007), in a study using laboratory Beagles and electronic training collars, compared three conditions: dogs that received a shock precisely when they engaged in a specific targeted behavior (allowing the shock to be associated with, and thereby predicted and to some extent controlled through, their own behavior), dogs that received a shock following a specific warning cue, and dogs that received shocks unpredictably, with no reliable cue or behavioral contingency. The dogs in the unpredictable condition showed the most pronounced and persistent physiological stress responses, including elevated cortisol, compared to the dogs for whom the stimulus was predictable and associated with their own behavior. This finding is broadly consistent with the uncontrollability/unpredictability framework: stressors that cannot be anticipated or influenced through behavior produce greater physiological disruption than stressors of identical physical intensity that can be predicted or influenced.

These findings should not be interpreted as evidence that the dogs in the unpredictable condition developed learned helplessness – the study did not assess generalized passivity or impaired escape learning in a subsequent novel context, the defining features of the classical phenomenon. Rather, the findings demonstrate that the core variables implicated in learned helplessness theory – predictability and controllability – have measurable, dog-specific welfare consequences independent of the physical intensity of the stimulus itself. It is this variable-level relevance, rather than any direct demonstration of learned helplessness as a phenomenon, that justifies the study's central place in this section.

2.3 Expectancy Models and Cognitive Interpretation

The original learned helplessness theory was, at its core, an expectancy model: organisms were proposed to form an expectation – that responses and outcomes are independent – and this expectation, once formed, generalized to new situations and governed subsequent behavior. This cognitive framing was influential because it offered an explanation for generalization: why an experience in one context (the harness) would affect behavior in an entirely different context (the shuttle box).

The 2016 reformulation does not abandon the role of expectation but relocates it. Rather than an expectation of helplessness being learned, what is learned (when learning occurs) is an expectation of control – and it is this latter expectation, encoded in prefrontal circuitry, that actively suppresses the default passivity response. In the absence of this learned expectation of control, the default response (passivity, mediated by the dorsal raphe nucleus) prevails. This is a substantively different cognitive architecture: rather than helplessness being the "added" cognitive state, control-detection is the added state, and its absence – not the presence of a helplessness schema – produces the passive phenotype.

2.4 Relationship to Prediction Error and Controllability

The prediction error framework, discussed in detail in the context of reinforcement learning (Schultz et al., 1997; see Prediction Error in Dogs: The Core Mechanism of Learning), offers a complementary perspective on controllability. Under conditions of control, an organism's actions reliably predict outcomes; prediction errors – the discrepancy between expected and actual outcomes – are informative and can be used to refine behavior. Under conditions of uncontrollability, no behavioral strategy reduces prediction error: outcomes remain unpredictable regardless of the organism's actions. This persistent, irreducible prediction error has been proposed as a contributing factor to the stress and motivational disruption associated with uncontrollable aversive events, though the relationship between prediction error signaling and the serotonergic/prefrontal circuitry described by Maier and Seligman (2016) has not been fully integrated into a single model.

For dogs, this suggests a further hypothesis, not yet directly tested: that training environments in which the dog's behavior reliably and predictably affects outcomes – regardless of whether those outcomes are rewarding or merely neutral – may support the prefrontal control-detection circuitry described in the 2016 reformulation, while environments in which outcomes occur independently of the dog's behavior (including, notably, environments where reinforcement is delivered inconsistently or where aversive events occur unpredictably) may fail to engage this circuitry, leaving the default passivity response unopposed.

3. Evidence in Dogs

3.1 Direct Studies in Dogs

Direct experimental studies of learned helplessness using dogs as subjects are, almost by definition, restricted to the historical literature of the 1960s and 1970s – the original Seligman, Maier, and colleagues' research program (Seligman & Maier, 1967; Seligman, Maier, & Geer, 1968; Overmier & Seligman, 1967) and a small number of related studies from the same period. These studies established the core phenomenon – that prior exposure to uncontrollable aversive stimulation interferes with subsequent escape/avoidance learning in a different context – using procedures that are not, and should not be, replicated in contemporary research or applied practice.

This creates a significant evidentiary situation for anyone seeking to apply learned helplessness theory to dogs in 2026: the foundational phenomenon was established in dogs, but the subsequent fifty years of theoretical refinement – including the entire 2016 neuroscientific reformulation – has been conducted almost exclusively in rodent models. There is, to this author's knowledge, no direct experimental test of the revised (2016) model in dogs. The applicability of the dorsal raphe nucleus / medial prefrontal cortex circuitry described by Maier and Seligman (2016) to dogs is, at present, an extrapolation from rodent neuroscience rather than a directly demonstrated finding.

3.2 Evidence from Studies of Aversive Training Methods

In the absence of direct learned helplessness research in dogs, the most relevant body of evidence comes from studies examining the behavioral and physiological effects of aversive training methods – particularly studies that manipulate or measure the controllability and predictability of aversive stimuli.

Schalke et al. (2007), introduced in Section 2.2, remains one of the most directly relevant studies. By comparing groups of dogs for whom an aversive electrical stimulus was associated with their own behavior (and therefore to some degree controllable and predictable) versus a group for whom the same stimulus occurred unpredictably and without behavioral contingency, the study provides dog-specific evidence that controllability and predictability influence physiological stress responses. The dogs in the unpredictable condition showed the most pronounced physiological stress responses. While this study did not assess the longer-term behavioral generalization effects that define learned helplessness in the classical sense – it examined acute physiological responses during training, not subsequent behavior in novel contexts – it provides evidence that the controllability/predictability distinction, central to learned helplessness theory, is behaviorally and physiologically meaningful for dogs.

Schilder and van der Borg (2004) examined dogs trained with shock collars and reported behavioral changes consistent with inhibited responding, including reduced responsiveness in contexts associated with training, persisting beyond the training sessions themselves. The European Society of Veterinary Clinical Ethology's position statement on electronic training devices explicitly identifies learned helplessness, alongside intensified fear and aggression, as a recognized risk of aversive training methods applied with poor timing or excessive intensity, citing the original Seligman framework (cited in the ESVCE position statement; see also Seligman, 1972).

It should be emphasized that these studies establish plausible mechanistic links between aversive training conditions and the variables (uncontrollability, unpredictability) known from the learned helplessness literature to produce behavioral and physiological disruption. They do not constitute direct demonstrations that dogs trained under such conditions develop learned helplessness in the full sense of the term – generalized passivity, motivational deficits, and impaired learning in subsequently novel contexts. This distinction matters for the integrity of the argument: the case for concern about aversive training methods rests on a convergence of related findings (controllability/predictability effects on stress physiology, behavioral inhibition, the general welfare literature on aversive methods) rather than on direct demonstrations of the classical learned helplessness phenomenon in companion dogs.

For a broader discussion of how aversive training methods affect dogs, see Aversive Training Methods: Neurological Effects in Dogs.

3.3 Evidence from Chronic Stress and Shelter Studies

A second relevant body of evidence comes from research on dogs in chronically stressful, low-control environments – particularly shelters, laboratory housing, and breeding facilities. Beerda et al. (1999a, 1999b) conducted a controlled study in which dogs were subjected to chronic social and spatial restriction and assessed across behavioral, hormonal, and immunological measures. The behavioral findings included increased low-posture behaviors, reduced behavioral variability, and signs consistent with chronic stress; the companion hormonal study documented changes in cortisol regulation consistent with sustained HPA axis activation.

Beerda et al. (2000) further characterized behavioral and hormonal indicators of what they termed "enduring environmental stress" in dogs, identifying a cluster of behaviors – including low body posture, reduced activity, and behavioral stereotypies – associated with prolonged exposure to environments offering little control over social contact, space, or stimulation.

While these studies do not use the term "learned helplessness" and were not designed within that theoretical framework, the conditions they examine – chronic, inescapable, low-control environments – are precisely the conditions identified by learned helplessness theory as producing the passivity and reduced behavioral variability described in both the original and revised formulations. The behavioral pattern described – reduced exploration, reduced behavioral variability, low-posture and withdrawal behaviors – is consistent with what would be predicted by learned helplessness theory applied to chronic environmental stress, though the studies themselves frame their findings in terms of chronic stress and coping rather than learned helplessness specifically.

3.4 Observable Behavioral Indicators

Drawing on the studies discussed above and on the broader stress and welfare literature, several behavioral indicators have been associated with states consistent with learned helplessness in dogs, though – as discussed further in Section 6 – none of these indicators is specific to learned helplessness, and all require interpretation in context.

Reduced response initiation. A reduction in the frequency of voluntary, exploratory, or operant behavior – the dog stops trying things. In a training or behavior modification context, this may present as a dog that no longer offers behaviors during shaping, that stops attempting to solve simple problems, or that shows minimal response to novel stimuli that would normally elicit investigation.

Reduced behavioral variability. A narrowing of the behavioral repertoire – the dog's responses become repetitive, predictable, and limited in range, even when the environment would support a wider range of responses. Beerda et al. (1999a, 2000) documented this pattern in chronically stressed dogs.

Low body posture and reduced activity. Persistent low, withdrawn body postures and reduced general activity levels, beyond what would be expected from simple rest or relaxation (see Section 5 for the distinction between this and genuine calm behavior).

Impaired new learning. Difficulty acquiring new operant contingencies, even straightforward ones, consistent with the "cognitive deficit" described in the original theory. This may present clinically as a dog that appears unable to learn new behaviors despite adequate motivation and clear contingencies, in contrast to dogs showing simple lack of motivation or attention problems.

Generalization across contexts. Perhaps the most theoretically significant indicator, though also the most difficult to observe directly: behavioral suppression that extends beyond the specific context in which the original uncontrollable experience occurred, appearing in contexts that are objectively safe and where the dog has not had any aversive experience.

3.5 Problems of Operationalization

The central operational problem in applying learned helplessness theory to dogs is that the theory was developed and validated using procedures – controlled exposure to inescapable aversive stimulation followed by tests of escape learning in a novel apparatus – that cannot ethically be replicated, and that have no direct analog in the everyday environments of companion, working, or shelter dogs. Real-world dogs are not exposed to a single, well-defined uncontrollable stressor followed by a clean test of subsequent learning; they experience complex, often chronic, multifactorial environments in which controllability varies across contexts, time, and specific stimuli.

This creates a fundamental measurement problem: how does one identify "learned helplessness" in a dog whose history involves an unknown combination of controllable and uncontrollable experiences, across multiple contexts, without a controlled baseline? The behavioral indicators listed in Section 3.4 are, individually, non-specific – each can result from multiple underlying states (see Section 6). The theoretical construct of learned helplessness, as rigorously defined in the experimental literature, may therefore be more useful as a framework for understanding why certain training conditions are harmful (the controllability/predictability analysis) than as a diagnostic category to be applied to individual dogs based on behavioral observation alone. This distinction – between learned helplessness as an explanatory framework versus a diagnosable condition – is revisited in Section 6.

4. Underlying Mechanisms

4.1 The HPA Axis and Cortisol

The hypothalamic-pituitary-adrenal (HPA) axis is the primary neuroendocrine stress response system and is central to the physiological signature of uncontrollable stress. Exposure to uncontrollable aversive events activates the HPA axis, resulting in the release of cortisol (in dogs, the primary glucocorticoid). Under conditions of control, HPA axis activation tends to be more limited and to resolve more quickly once the stressor is terminated or escaped. Under conditions of uncontrollability, HPA axis activation tends to be more pronounced and more persistent (Maier & Seligman, 2016).

In dogs, elevated and dysregulated cortisol has been documented in chronically stressed populations, including shelter dogs (Beerda et al., 1999b) and dogs trained with poorly timed aversive stimuli (Schalke et al., 2007). Of particular note is the finding, in some chronic stress research, of reduced cortisol reactivity alongside elevated baseline levels – a pattern consistent with a dysregulated, rather than simply hyperactive, HPA axis, and one that may reflect long-term adaptation to chronic stress exposure (discussed in the context of extreme life history dogs in shelter populations).

For a detailed discussion of HPA axis function and chronic stress in dogs, see The Neurobiology of Chronic Stress in Dogs: Cortisol Impact.

4.2 Serotonergic Systems: The Dorsal Raphe Nucleus

The central neurobiological claim of the 2016 reformulation concerns the dorsal raphe nucleus (DRN), a brainstem structure that is the primary source of serotonergic projections throughout the brain. In the rodent model developed by Maier and colleagues, exposure to inescapable, uncontrollable stress produces sustained activation of serotonergic neurons in the DRN, and this sustained activation is causally linked to the behavioral passivity that defines learned helplessness. Pharmacological or optogenetic inhibition of DRN serotonergic activity during stress exposure prevents the development of the passive phenotype, while activation of these neurons in the absence of stress can produce passivity-like effects (Maier & Seligman, 2016; Amat et al., 2010).

This places serotonergic signaling at the center of the mechanistic account – a notable departure from earlier dopamine-centric accounts of motivation and learned helplessness. Direct evidence regarding DRN function in dogs under conditions of controllable versus uncontrollable stress does not exist; the relevance of this circuitry to dogs is, at present, inferred from the conserved nature of brainstem monoaminergic systems across mammals rather than directly demonstrated.

4.3 Dopaminergic Systems and Motivation

While the 2016 reformulation emphasizes serotonergic mechanisms for the passivity response itself, dopaminergic systems remain centrally relevant to the motivational deficits associated with learned helplessness – the reduced initiation of voluntary, goal-directed behavior. As discussed in the context of reinforcement learning (Section 2.4; see also Dopamine and Learning in Canine Neurochemistry), dopaminergic signaling in mesolimbic circuits is closely tied to the initiation and maintenance of approach behavior and to the encoding of action-outcome contingencies.

Under conditions of uncontrollability, where actions do not reliably predict outcomes, the informational value of dopaminergic prediction error signals is degraded – there is no consistent contingency for the system to learn. Whether chronic exposure to such conditions produces lasting changes in dopaminergic system function in dogs, analogous to findings in rodent models of chronic stress and anhedonia, has not been directly investigated. The motivational deficit described in learned helplessness theory – reduced initiation of voluntary behavior – is consistent with reduced dopaminergic drive, but the causal relationship between uncontrollable stress, serotonergic activation in the DRN, and downstream dopaminergic function remains to be fully characterized even in the rodent literature, let alone in dogs.

4.4 Prefrontal-Amygdala Dynamics and the Detection of Control

The medial prefrontal cortex (mPFC) occupies the central role in the 2016 reformulation as the site where the detection of control is computed and where, when control is detected, inhibitory output to the dorsal raphe nucleus is generated – actively suppressing the default passivity response. This places the mPFC in a role analogous to, but mechanistically distinct from, its role in fear extinction (where the infralimbic cortex inhibits amygdala-mediated fear responses) and in impulse control more broadly.

The amygdala remains relevant to learned helplessness as a structure implicated in the processing of aversive stimuli and in the generation of fear and anxiety responses that frequently co-occur with, and may be difficult to distinguish from, learned helplessness (see Section 6). The relationship between mPFC-mediated control detection and amygdala-mediated fear processing is an area of active investigation in the rodent literature, with the general picture being one of mPFC regions exerting top-down regulatory influence over both amygdala and DRN function when control has been learned and detected.

For dogs, the prefrontal cortex has been discussed extensively in the context of self-control and impulse regulation (see Prefrontal Cortex and Self-Control in Dogs), and the broader prefrontal-amygdala circuitry has been discussed in the context of reactivity (see Reactivity in Dogs: A Neurological Perspective). The specific application of the control-detection circuitry described by Maier and Seligman (2016) to these existing frameworks for canine prefrontal function represents a promising but currently unexplored integration.

4.5 Stress-Induced Changes in Learning Processes

A consistent theme across the learned helplessness literature, present in both the original and revised formulations, is that uncontrollable stress does not merely produce passivity in the moment – it changes how subsequent learning occurs. The "cognitive deficit" of the original theory and the control-detection failure of the revised theory both describe, in different terms, an impairment in the acquisition of new response-outcome contingencies following exposure to uncontrollable stress.

Mechanistically, this may involve several converging processes: elevated cortisol has well-documented effects on hippocampal function and on the consolidation of new learning; sustained serotonergic activation in the DRN, per the 2016 model, directly promotes passivity and may reduce the behavioral output through which new contingencies could be discovered; and chronic stress more broadly has been associated with reduced behavioral flexibility and increased reliance on previously established (often maladaptive) response patterns. For dogs in behavior modification contexts, this convergence has a clear practical implication: a dog whose behavioral history includes substantial exposure to uncontrollable or unpredictable aversive events may plausibly show impaired acquisition of new behaviors during early treatment, independent of the specific training methods used – a possibility that should inform expectations about the pace of progress and that may require addressing the stress and control variables themselves before, or alongside, direct behavioral training.

5. Learned Helplessness vs. Calm Behavior

This section addresses a problem of particular practical importance, frequently encountered by trainers, behavior therapists, and owners: the difficulty of distinguishing genuine calm, relaxed behavior from behavioral suppression that superficially resembles it.

5.1 The Interpretation Problem

A dog can present, to an observer, as calm: lying still, minimal movement, reduced vocalization, no apparent conflict signals, no attempts to leave or avoid the situation. This presentation can result from at least two very different underlying states. In the first, the dog is genuinely relaxed – its physiological arousal is low, it could engage with the environment if motivated to but is not currently motivated to, and it would readily resume normal behavioral variability if the context changed. In the second, the dog's behavioral output has been suppressed – through chronic stress, repeated aversive experience, or the kind of uncontrollable/unpredictable conditions discussed in Sections 2–4 – such that the absence of behavior reflects an inability or strong disinclination to respond, rather than a positive absence of motivation to respond.

These two states can be behaviorally almost indistinguishable using simple observation, particularly to an untrained observer, and even to experienced observers without additional information about the dog's history and physiological state. This is not a minor academic distinction. A dog presenting the second pattern – behavioral suppression – may be assessed by an owner or even a professional as "calm," "settled," or "well-behaved," when the underlying state is one of chronic stress or a learned-helplessness-consistent pattern of reduced responding. The practical consequences of this misinterpretation can be substantial: a dog assessed as calm is unlikely to be identified as needing intervention, and behavioral suppression may be inadvertently reinforced or even deliberately trained as a desirable outcome.

5.2 Indicators That May Help Distinguish the Two States

While no single indicator is definitive, several lines of evidence may help differentiate genuine calm behavior from suppression-based stillness.

Responsiveness to positive stimuli. A genuinely relaxed dog typically remains responsive to the introduction of positive stimuli – food, play invitations, social engagement – showing an appropriate increase in arousal and approach behavior. A dog in a suppressed state may show reduced or absent responses even to stimuli that would normally be highly motivating, a pattern sometimes described as anhedonia-like in the broader stress literature.

Ease of behavioral transition. A genuinely relaxed dog typically transitions easily between states – from rest to alert investigation to play and back to rest – in response to environmental changes. A dog in a suppressed state may show a "stuck" quality, remaining in a low-activity state even when the environment changes in ways that would normally elicit a behavioral response.

Physiological indicators. Where available, physiological measures – heart rate variability, cortisol – can help distinguish low arousal associated with genuine relaxation from the dysregulated patterns associated with chronic stress (Beerda et al., 1999b, 2000). These measures are not generally available in everyday practice but may be relevant in clinical assessment.

Postural quality. Beerda et al. (1999a, 2000) and related work describe specific postural patterns – low body posture, tucked positioning, reduced muscle tone in ways inconsistent with relaxed rest – associated with chronic stress, distinguishable on close observation from the loose, often asymmetric postures of genuine relaxation (e.g., a dog lying fully on its side with loose limbs versus a dog curled tightly with tense musculature).

History and context. Perhaps most importantly, the dog's history and the context in which the calm presentation occurs provide critical interpretive information. A dog that has just undergone an aversive experience, that is in an environment it has previously found aversive, or that has a documented history of chronic stress exposure warrants more cautious interpretation of an apparently calm presentation than a dog with no such history resting in a familiar, positive environment.

5.3 Why This Matters for Training and Welfare Assessment

The conflation of behavioral suppression with calm behavior has direct practical consequences across multiple contexts discussed elsewhere in this article and in related literature.

In aversive training contexts, a reduction in problem behavior following the application of an aversive stimulus may be interpreted as successful training – the dog has "calmed down." Section 3.2 discussed evidence that aversive methods, particularly those involving unpredictable or poorly timed stimuli, are associated with behavioral inhibition and stress indicators. A reduction in observable behavior under these conditions should not be assumed to indicate improved welfare or successful resolution of the underlying issue; it may instead indicate exactly the kind of behavioral suppression discussed in this section.

In shelter and rehoming contexts, a dog that presents as quiet and undemanding may be assessed as well-adjusted, when this presentation may instead reflect the chronic stress patterns documented by Beerda et al. (1999a, 1999b, 2000) in dogs subjected to prolonged social and spatial restriction.

In behavior modification contexts more broadly, the goal of treatment is sometimes implicitly or explicitly framed as "reducing problem behavior" – and a dog that has stopped exhibiting a problem behavior (such as reactive barking or lunging) may be considered improved, regardless of whether the underlying emotional state has changed. This conflation is discussed extensively in the broader literature on the distinction between behavior and emotion (see Behavior Does Not Equal Emotion: Limits of Inferring Internal States) and is directly relevant to the learned helplessness framework: a dog that has "stopped" showing a behavior because it has learned that responding has no effect on the outcome (the core learned helplessness mechanism, in either its original or revised form) has not had its underlying problem resolved – it has, in a meaningful sense, given up.

The practical implication is that assessments of training success or welfare status that rely solely on the presence or absence of observable behavior are insufficient. Assessment should additionally consider the dog's responsiveness, behavioral flexibility, and – where the history is known – the conditions under which the apparent calm was achieved.

6. Research Gaps and Controversies

6.1 The Difficulty of Differential Diagnosis

The behavioral indicators associated with learned helplessness (Section 3.4) substantially overlap with the indicators of several other states, and distinguishing between them is one of the most significant unresolved challenges in applying this framework to dogs.

Fear involves heightened arousal, avoidance, and escape attempts directed at a specific or generalized threat. In its acute form, fear is behaviorally quite distinct from learned helplessness – fear typically involves increased, not decreased, behavioral output (avoidance, escape attempts, vigilance). However, chronic or overwhelming fear, particularly when escape is not possible, can transition into the kind of behavioral shutdown that overlaps substantially with learned helplessness presentations. The relationship between chronic fear and learned-helplessness-consistent passivity is an area where the 2016 reformulation's framing – passivity as a default response to overwhelming aversive conditions, with fear/anxiety-like qualities – may be particularly useful, as it does not require treating fear and helplessness as entirely separate categories.

Anxiety, characterized by anticipatory arousal in the absence of an immediate, identifiable threat, is also closely linked to uncontrollability in the broader literature – anxiety has been characterized as, in part, a response to unpredictability and lack of control over future events. The overlap between anxiety and learned helplessness at the level of underlying neurobiology (both implicate the amygdala, HPA axis dysregulation, and prefrontal regulatory failure) is substantial, and behavioral presentations may be difficult to distinguish without detailed contextual information.

Behavioral inhibition, a temperament-level construct describing a stable individual tendency toward caution, withdrawal, and reduced exploration in novel situations, represents a potential confound of a different kind: a dog with a temperamentally high level of behavioral inhibition may present with reduced exploratory behavior and behavioral variability for reasons unrelated to learned helplessness – this may simply be how the individual dog responds to novelty, independent of any history of uncontrollable aversive experience.

Freeze responses, a component of the broader fight-flight-freeze-fawn repertoire, represent acute, often brief, immobility responses to perceived threat. Freeze is generally understood as an active defensive response (often associated with heightened, not reduced, physiological arousal) rather than the more chronic, lower-arousal passivity associated with learned helplessness, though both can present, superficially, as "the dog isn't moving."

Shutdown, a term widely used in applied dog behavior contexts but with less precise definition in the empirical literature, generally refers to a state of pronounced behavioral suppression, reduced responsiveness, and apparent disengagement from the environment, often following overwhelming or prolonged aversive experience. Of all the related constructs, shutdown as commonly described in applied contexts has perhaps the closest descriptive overlap with learned helplessness, but the term is used inconsistently across sources and has not been operationalized with the rigor of the laboratory learned helplessness literature.

Depression-like states in non-human animals, a construct with its own substantial methodological controversies even in well-studied laboratory species, share with learned helplessness the core features of reduced motivation, reduced responsiveness to reward (anhedonia-like presentations), and behavioral passivity. Indeed, learned helplessness paradigms have historically been used as an animal model of depression in pharmacological research (Maier & Seligman, 2016). Whether "depression" is a useful or appropriate construct to apply to dogs, and how it would be distinguished from learned helplessness if so, remains an open question without a clear answer in the current literature.

The overarching point is that these constructs are not cleanly separable categories with distinct behavioral signatures. They overlap substantially, both at the level of behavioral presentation and at the level of underlying neurobiology (HPA axis, amygdala, prefrontal cortex, serotonergic and dopaminergic systems all appear, in varying configurations, across all of these constructs). A rigorous differential diagnosis in an individual dog would require detailed history, longitudinal observation, and ideally physiological assessment – resources not typically available in applied practice. This does not mean the constructs are useless; it means that practical application should favor frameworks (such as the controllability/predictability analysis) that are robust to this overlap, rather than diagnostic labels that imply a precision the field cannot currently support.

6.2 The Scarcity of Direct Experimental Studies in Companion Dogs

As discussed in Section 3.1, the foundational learned helplessness literature in dogs is restricted to historical studies using procedures that cannot be ethically replicated, and the 2016 neuroscientific reformulation has not been directly tested in dogs at all. This creates a peculiar situation: dogs hold a historically central place in the development of learned helplessness theory, yet the theory's most current and mechanistically detailed form rests on evidence from a different species (primarily rats).

This is not merely a gap to be filled by more research using the original aversive paradigms – such research would be neither ethical nor necessary. Rather, the gap calls for research using non-invasive, welfare-compatible designs that can examine the controllability/predictability variables central to the theory without exposing dogs to the kind of procedures used in the original studies. Some existing research – Schalke et al. (2007), the chronic stress literature (Beerda et al., 1999a, 1999b, 2000), and studies of shelter and rehoming stress – approaches this from adjacent angles, but a more direct, theoretically integrated research program examining controllability, predictability, and their relationship to the specific neurobiological mechanisms described in the 2016 reformulation, using welfare-compatible methods, does not currently exist for dogs.

6.3 Generalization from Rodent Neuroscience

The entire mechanistic apparatus of the 2016 reformulation – DRN-mediated default passivity, mPFC-mediated control detection, the specific circuitry connecting these structures – derives from rodent research, primarily using rats. The general principle that basic mammalian stress and motivational circuitry is conserved across species provides some basis for expecting broad applicability to dogs, but the specific parameters – the relative balance of serotonergic and dopaminergic contributions, the precise role of prefrontal subregions, individual and breed differences in baseline reactivity of these systems – are unknown for dogs. Given the substantial differences in domestication history, social cognition, and human-directed behavior between dogs and rats (discussed extensively in the social learning literature; see Social Learning in Dogs: Learning Through Observation and Demonstration), caution is warranted in assuming that findings from rodent models transfer without modification to dogs, particularly regarding any role that the dog-human relationship and dog-human communication might play in either producing or ameliorating learned-helplessness-consistent states – a dimension entirely absent from rodent models.

7. Practical Applications

7.1 Why Predictability Matters

The evidence reviewed in this article converges on predictability as a key variable affecting welfare independent of, though related to, controllability. Schalke et al. (2007) demonstrated that unpredictable aversive stimuli produce greater physiological stress responses than predictable ones of identical intensity. The broader uncontrollability literature suggests that unpredictability contributes to chronic stress and may interfere with the development of the control-detection mechanisms described in the 2016 reformulation.

For training and management, this has direct implications. Environments, routines, and interactions that are predictable – where the dog can reliably anticipate what will happen, when, and in response to what – are likely to support lower baseline stress and better conditions for the kind of learning (control detection) that the 2016 model identifies as protective. This applies not only to the presence or absence of aversive events but to the broader structure of the dog's environment: consistent routines, predictable handler behavior, and clear, consistent cues all contribute to an environment in which the dog can form reliable expectations.

This does not mean that novelty or variation should be eliminated – indeed, excessive rigidity in environment and routine may itself constrain the behavioral variability associated with good welfare. The relevant distinction is between novelty that the dog can engage with and learn about (supporting exploration and control-detection) versus unpredictability in aversive or high-stakes contexts specifically, where the inability to anticipate outcomes is most directly linked to the mechanisms described in this article.

7.2 Agency as a Protective Factor

The concept of agency – an individual's experience of being a causal agent capable of producing effects in its environment – has emerged across several lines of evidence discussed in this article as a unifying theme, and it deserves treatment as a concept in its own right rather than as a byproduct of other discussions. Agency can be understood as the practical, experiential counterpart to what the 2016 reformulation describes mechanistically as control detection: agency is what an organism has when its prefrontal control-detection circuitry is actively engaged and signaling that its behavior matters.

Choice. The opportunity to make meaningful choices – which path to take on a walk, whether to approach or retreat from a novel stimulus, which of several available activities to engage in – provides direct, repeated experiences of behavior-outcome contingency. Each choice that produces a perceptible consequence is, in the terms of the 2016 model, a small instance of control detection. Training and management approaches that build in genuine choice points, rather than specifying every aspect of the dog's behavior, may therefore serve a function beyond simple enrichment: they may provide the recurring experiences of efficacy that the control-detection circuitry requires to develop and remain engaged.

Control over aversive or uncomfortable experiences. Beyond positive choices, the ability to terminate, escape, or otherwise influence uncomfortable or aversive experiences – a loud noise, an unwanted approach from another dog, handling that the dog finds uncomfortable – is directly the variable that the original learned helplessness research identified as protective. Cooperative care approaches in veterinary and grooming contexts, in which dogs are trained to offer behaviors that start or stop a procedure, operationalize exactly this principle: the dog's behavior determines whether an aversive or uncomfortable event continues, restoring the behavior-outcome contingency that uncontrollable aversive events remove.

Behavioral immunization. The rodent findings on behavioral immunization (Amat et al., 2010) – in which prior experience of control over one stressor reduces the behavioral and neurochemical impact of a later, unrelated stressor – suggest that the protective effects of agency may not be narrowly tied to the specific context in which control was experienced. If this generalizes to dogs, it would imply that building agency into a dog's everyday experiences – in contexts that have nothing to do with a specific behavioral problem – could have protective value for the dog's responses to future, unrelated stressors. This remains a hypothesis for dogs rather than a demonstrated finding, but it is a hypothesis with substantial theoretical grounding and considerable practical appeal, since it suggests that agency-building interventions need not be narrowly targeted to be valuable.

Enrichment as agency, not just stimulation. Environmental enrichment is often framed primarily in terms of stimulation – providing novel objects, scents, or activities to occupy the dog. The agency framework suggests an additional dimension: enrichment that allows the dog to act upon the environment and observe the effects of its actions (puzzle feeders, scent work, shaping-based games, choice-based environmental setups) may be more relevant to the mechanisms discussed in this article than enrichment that is purely passive or observational, even if both are subjectively engaging to the dog.

Together, these threads position agency as the practical, applied counterpart to the theoretical construct of control detection: where learned helplessness theory describes what happens when behavior-outcome contingencies are absent, agency-focused practice describes how to ensure that dogs have ongoing, accessible experiences of behavior-outcome contingency across multiple domains of their lives. If the 2016 reformulation is correct that this detection is the actively learned, protective component of the system – with passivity as the default in its absence – then agency is not an optional welfare enhancement but addresses a foundational requirement of the system itself.

7.3 Implications for Training and Behavior Therapy

Several practical implications follow from the framework developed in this article, while maintaining appropriate caution about the strength of the underlying evidence in dogs specifically.

Avoid training methods that rely on unpredictable or uncontrollable aversive stimuli. This is the most direct implication of the controllability/predictability framework and is consistent with the broader welfare evidence on aversive training methods (Section 3.2). The relationship between punishment and learned helplessness, and more broadly between aversive training and learned helplessness, is not that punishment automatically produces helplessness – it is that methods in which aversive consequences occur with poor timing, inconsistent application, or without a clear behavioral contingency the dog can learn are precisely the conditions identified by learned helplessness research as most likely to produce stress and behavioral disruption.

Be cautious about interpreting reduced problem behavior as success. As discussed in Section 5, a reduction in observable behavior is not equivalent to improved welfare or resolved underlying issues. Assessment of training and behavior modification outcomes should include indicators of behavioral flexibility, responsiveness to positive stimuli, and willingness to engage – not only the absence of unwanted behavior.

Build histories of successful, dog-driven problem-solving. Consistent with the original "cure" findings of Seligman, Maier, and Geer (1968) – in which recovery from learned helplessness required the dog's own experience of producing a response that worked, not merely the removal of the aversive stimulus – behavior modification programs may benefit from deliberately incorporating opportunities for the dog to discover that its own behavior produces effects. Shaping-based training, in which the dog's voluntary behavior is directly and consistently linked to outcomes, may serve this function particularly well, independent of the specific behaviors being trained.

Consider environmental predictability as a foundational welfare variable, not only in formal training contexts but in the dog's overall daily environment – routines, handler consistency, and the structure of the living environment.

In behavior modification for dogs with suspected histories of chronic uncontrollable stress (e.g., dogs from neglect, hoarding, or poorly run shelter or breeding environments), expectations about the pace of new learning should account for the possibility of plausibly impaired acquisition (Section 4.5), and initial intervention may need to focus on building basic experiences of control and predictability before – or alongside – addressing specific target behaviors.

For further discussion of building agency and choice into training and behavior modification, see Arousal Regulation in Dogs: Neurophysiology and Training and The Neurobiology of Frustration in Dogs.

8. Conclusion

Learned helplessness occupies a unique position in the history of canine behavioral science: the phenomenon was discovered in dogs, but its most current theoretical formulation rests almost entirely on evidence from other species. The 2016 reformulation by Maier and Seligman represents a substantial departure from the original 1967 theory – not a minor refinement but, in the authors' own framing, a reversal of the original causal account. Passivity in the face of overwhelming, uncontrollable aversive events is now understood as the default mammalian response, mediated by serotonergic activity in the dorsal raphe nucleus, while active coping requires learned, prefrontally-mediated control detection that actively inhibits this default.

For dogs, the direct evidentiary basis for this revised framework is, at present, absent. What exists is a body of related, dog-specific evidence – on the effects of unpredictable versus predictable aversive stimuli (Schalke et al., 2007), on the behavioral and hormonal consequences of chronic, low-control environments (Beerda et al., 1999a, 1999b, 2000), and on the broader welfare effects of aversive training methods – that is consistent with, and can be productively organized by, the controllability/predictability framework central to learned helplessness theory, without constituting direct tests of the revised mechanistic model.

The differential diagnosis problem discussed in Section 6 is not a peripheral concern but central to the responsible use of this framework. The behavioral indicators associated with learned helplessness overlap substantially with fear, anxiety, behavioral inhibition, freeze responses, shutdown, and depression-like states, and a rigorous distinction between these states in an individual dog is rarely possible using behavioral observation alone. The most defensible practical use of learned helplessness theory for dogs is therefore not as a diagnostic label to be applied to individual animals, but as an explanatory framework that identifies controllability and predictability as central welfare variables – variables that can be assessed and modified in training and management regardless of whether a specific dog's presentation can be confidently attributed to "learned helplessness" as opposed to a related state.

The distinction between genuine calm behavior and behavioral suppression, addressed in Section 5, may be the most practically consequential single point in this article. A dog that has stopped struggling has not necessarily been helped. The responsible application of learned helplessness theory to dogs requires holding this possibility in mind – not as a source of alarm about every quiet dog, but as a corrective to assessment practices that equate the absence of observable behavior with the presence of wellbeing.

Key Insights (Takeaways)

1. Learned helplessness theory has undergone a fundamental reformulation since its original 1967 description in dogs. The 2016 revision (Maier & Seligman) proposes that passivity to overwhelming aversive events is the default, unlearned mammalian response (mediated by the dorsal raphe nucleus), and that what requires learning is the detection of control, mediated by the medial prefrontal cortex, which actively inhibits this default. This reformulation has not been directly tested in dogs but, if applicable, substantially reframes how the theory should be applied.

2. Direct experimental evidence for learned helplessness in dogs is restricted to historical studies using procedures that cannot and should not be replicated. The relevant contemporary evidence for dogs comes from adjacent research – on unpredictable versus predictable aversive stimuli (Schalke et al., 2007) and on chronic, low-control environments (Beerda et al., 1999a, 1999b, 2000) – which is consistent with, but does not directly test, the learned helplessness framework, particularly its revised neuroscientific form.

3. Learned helplessness overlaps substantially, both behaviorally and neurobiologically, with fear, anxiety, behavioral inhibition, freeze responses, shutdown, and depression-like states. A confident differential diagnosis in an individual dog using behavioral observation alone is rarely possible. The framework is most defensibly used to identify controllability and predictability as welfare-relevant variables, rather than as a diagnostic category.

4. The distinction between genuine calm behavior and behavioral suppression is of major practical importance. A dog that has stopped exhibiting problem behavior, or that presents as quiet and undemanding, may be in a state of chronic stress or learned-helplessness-consistent passivity rather than relaxation. Assessment should consider responsiveness to positive stimuli, ease of behavioral transition, and the dog's history and context – not only the absence of observable behavior.

5. Practical application centers on predictability and control as foundational welfare variables: avoiding training methods involving unpredictable or uncontrollable aversive stimuli, building histories of dog-driven successful problem-solving, and maintaining cautious expectations about learning pace in dogs with suspected histories of chronic uncontrollable stress. These recommendations are grounded in the controllability/predictability framework rather than in direct dog-specific tests of the 2016 reformulation, and should be applied with this evidentiary status in mind.

References

Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychology, 87(1), 49–74. https://doi.org/10.1037/0021-843X.87.1.49

Amat, J., Aleksejev, R. M., Paul, E., Watkins, L. R., & Maier, S. F. (2010). Behavioral control over shock blocks behavioral and neurochemical effects of later social defeat. Neuroscience, 165(4), 1031–1038. https://doi.org/10.1016/j.neuroscience.2009.11.005

Beerda, B., Schilder, M. B. H., Van Hooff, J. A. R. A. M., De Vries, H. W., & Mol, J. A. (1999a). Chronic stress in dogs subjected to social and spatial restriction. I. Behavioral responses. Physiology & Behavior, 66(2), 233–242. https://doi.org/10.1016/S0031-9384(98)00289-3

Beerda, B., Schilder, M. B. H., Bernadina, W., Van Hooff, J. A. R. A. M., De Vries, H. W., & Mol, J. A. (1999b). Chronic stress in dogs subjected to social and spatial restriction. II. Hormonal and immunological responses. Physiology & Behavior, 66(2), 243–254. https://doi.org/10.1016/S0031-9384(98)00290-X

Beerda, B., Schilder, M. B. H., Van Hooff, J. A. R. A. M., De Vries, H. W., & Mol, J. A. (2000). Behavioural and hormonal indicators of enduring environmental stress in dogs. Animal Welfare, 9(1), 49–62. https://doi.org/10.1017/S0962728600022247

Maier, S. F., & Seligman, M. E. P. (2016). Learned helplessness at fifty: Insights from neuroscience. Psychological Review, 123(4), 349–367. https://doi.org/10.1037/rev0000033

Overmier, J. B., & Seligman, M. E. P. (1967). Effects of inescapable shock upon subsequent escape and avoidance responding. Journal of Comparative and Physiological Psychology, 63(1), 28–33. https://doi.org/10.1037/h0024166

Schalke, E., Stichnoth, J., Ott, S., & Jones-Baade, R. (2007). Clinical signs caused by the use of electric training collars on dogs in everyday life situations. Applied Animal Behaviour Science, 105(4), 369–380. https://doi.org/10.1016/j.applanim.2006.11.002

Schilder, M. B. H., & van der Borg, J. A. M. (2004). Training dogs with help of the shock collar: Short and long term behavioural effects. Applied Animal Behaviour Science, 85(3–4), 319–334. https://doi.org/10.1016/j.applanim.2003.10.004

Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate of prediction and reward. Science, 275(5306), 1593–1599. https://doi.org/10.1126/science.275.5306.1593

Seligman, M. E. P. (1972). Learned helplessness. Annual Review of Medicine, 23(1), 407–412. https://doi.org/10.1146/annurev.me.23.020172.002203

Seligman, M. E. P., & Maier, S. F. (1967). Failure to escape traumatic shock. Journal of Experimental Psychology, 74(1), 1–9. https://doi.org/10.1037/h0024514

Seligman, M. E. P., Maier, S. F., & Geer, J. H. (1968). The alleviation of learned helplessness in the dog. Journal of Abnormal Psychology, 73(3), 256–262. https://doi.org/10.1037/h0025831