International Canine Legislation Action Committee
Evidence-Based Position Paper of Training Methods, Canine Welfare and the Use of Aversive Techniques. (v3.0)
The Science of Dog Training:
Evidence-Based Position Paper of Training Methods, Canine Welfare and the Use of Aversive Techniques
Tressie Dutchyn, (Hon)BA, MA
Animal Advocate
Will Bangura, M.S., CAB-ICB, CBCC-KA, CPDT-KA, FFCP
Certified Canine Behaviorist
INTERNATIONAL CANINE LEGISLATION ACTION COMMITTEE (ICLAC)
August 2026
This evidenced-based position paper integrates over a century of evolutionary biology, behavioural science, and contemporary veterinary medicine to advocate for the systemic eradication of electronic shock collars (e-collars) and secondary aversive training tools and methods.
The foundational principles of animal behaviour spanning Charles Darwin's evolutionary insights, Ivan Pavlov's classical conditioning, B.F. Skinner’s operant learning, and the developmental timelines established by Scott and Fuller have been systematically validated and expanded by modern cognitive, neurological, and welfare research. A vast body of modern literature including comprehensive scientific reviews by researchers such as Albuquerque, Andics, Berns, Buttner, Call, Casey, Fugazza, Hare, Kaminski, Miklósi, Morrill, Salomons, Topál and Ziv - confirms that empirical data has firmly replaced historical folklore. These studies prove that dogs are predictive learners, social partners, and emotional beings whose co-evolutionary partnership with humans extends back over 14,000 years, as documented in landmark genetic and archaeological timelines published by Bergström, Janssens, and Perri.
Rather than a passive or "soft" approach, force-free, reward-based training is structurally verified as the most effective standard for animal learning and canine psychological health.
· Global Veterinary Consensus Opposing Aversive Training Methods: Over the past decade, a major paradigm shift has occurred within the global veterinary community regarding animal training and behaviour modification. Historically, various methodologies ranging from positive reinforcement to physical punishment were debated. However, contemporary guidelines published by leading behavioural colleges, international networks, and national veterinary medical associations reveal an absolute, evidence-based consensus: the use of aversive training techniques is universally discouraged, and in several instances, explicitly condemned due to severe animal welfare risks.
· Humane Society Consensus Opposing Aversive Training Methods: International animal welfare organizations and humane societies maintain a unified, evidence-based stance against the use of aversive training methods, devices, and philosophies. These organizations operate under the mandate that humane animal care must minimize stress, fear, anxiety, and pain. While historically some training methodologies incorporated punishment or coercion to alter behaviour, modern policies across major global humane entities uniformly advocate for exclusive adherence to reward-based, force-free training protocols.
· Welfare Risks: Peer-reviewed data synthesized by researchers such as Herron establishes that suppressing surface symptoms through physical or electrical coercion creates severe secondary complications, acute situational anxiety, and chronic physiological harm.
· Efficacy Disparity: Controlled behavioural studies led by China demonstrate that alternative methods have no training advantage. Conversely, clinical trials by Hiby, Rooney, Cowan, and Vieira de Castro prove that relationship-based frameworks prioritizing positive reinforcement yield superior task willingness, stronger attachment security, and lower stress metrics.
The real-world viability of a force-free paradigm is robustly demonstrated by successful regulatory frameworks across several international jurisdictions:
· The United Kingdom (Wales): Outlawed the active use of all electronic shock devices under the Animal Welfare (Electronic Collars) (Wales) Regulations 2010, executed pursuant to the Animal Welfare Act 2006.
· Austria: Austria’s Federal Act on the Protection of Animals (enacted in 2004) established a strict zero-tolerance rule against aversive dog training tools, banning their use, sale, and possession.
· France: The Assemblée Nationale (2023) mobilized to phase out abusive training devices, shifting standard pet husbandry toward positive reinforcement. However, the absolute restriction on professional use was fully realized when the Ministère de l'Agriculture et de la Souveraineté alimentaire (2025) executed an updated ministerial decree strictly governing animal health and safety protections.
· Germany: Prohibits devices utilizing direct electrical currents under Section 3, Number 11 of the Tierschutzgesetz (Animal Welfare Act), a mandate upheld by the German Federal Administrative Court, which ruled that proof of individual physical injury is not required to enforce the ban.
· The Netherlands: Banished shock-delivering tools via an amendment to Article 1.3 of the Besluit houders van dieren (Animal Keeper Decree), building upon a preceding structural ban targeting sharp prong collars.
· Spain: Article 27 of Law 7/2023 explicitly outlaws aggressive handling devices. It is strictly illegal to use electric shock collars, impulse collars, spike/prong collars, or choking mechanisms that can inflict physical or psychological injury on companion animals.
· The Nordic Region: Maintained an absolute regional consensus criminalizing aversive training technology under the national animal welfare acts of Denmark, Norway, Sweden, Finland, and Iceland. As validated by the Finnish Kennel Club and global pet technology reviews compiled by Satellai, these frameworks restrict lawful gear strictly to passive GPS locators.
· Slovenia: Slovenia stands as a prominent European nation in the systemic push against coercive training practices. The country officially banned the use of electronic shock collars in 2011 as part of a series of progressive, sweeping reforms to its national Animal Welfare Act (Zakon o zaščiti živali).
The uncompromised performance of companion, sport, and working assistance dogs within these restricted nations provides real-world evidence that aversive equipment is entirely unnecessary. To transform the human-canine bond into a foundational source of community safety as envisioned in social bonding studies by Cook and Topál, this paper urges legislators to enact the following policy directives:
1. Importation Controls: Implement restrictive federal border customs regulations to prevent the entry of electronic shock devices and mechanical prong components.
2. Statutory Harmonization: Amend local animal protection acts to establish unified, nationwide bans on the sale, commercial distribution, public possession, and active deployment of all aversive training tools and methods.
1. Introduction: Why Science Matters for How We Live with Dogs
2. The Early Foundations: Darwin, Observation and Pavlov
2.1 Darwin and the Comparative View
2.2 Pavlov and Classical Conditioning
3. Behaviourism, Learning Theory and Dogs in the Laboratory
4. Developmental Science: Critical Periods and the Scott and Fuller Studies
5. The Cognitive Revolution: How Dogs Think and Communicate
5.1 The Family Dog Project and Duke Canine Cognition Center
5.2 The Emory Dog Project
6. Domestication, Genetics and Individual Differences
7. Learning Theory Applied: How Dogs Actually Learn
8. The Welfare Evidence: Aversive Methods and Their Costs
8.1 Stress Physiology and Behaviour During Training
8.2 Mood State Outside the Training Context
8.3 Behaviour Problems and the Human and Dog Relationship
8.4 Efficacy
8.5 Professional Positions
9. Recognition of Sentience and the Evolution of Welfare Law
10. What We Now Know About Dogs
11. Why Reward-Based Training
11.1 Effectiveness
11.2 Welfare
11.3 Relationship
11.4 Generality and Safety
11.5 Alignment with How Dogs Learn
12. Limitations and Strength of the Evidence
13. Examination of Contrary Claims
14. Science, Ethics and Putting Knowledge into Practice
14.1 Putting the Science into Practice
14.2 Evidence from Jurisdictions That Have Restricted Aversive Tools
15. Conclusion
References
1. Introduction: Why Science Matters for How We Live with Dogs
Dogs have lived alongside humans for a very long time. The oldest remains that can be attributed to a domestic dog, without serious dispute, come from the Bonn-Oberkassel burial site in Germany, where a dog of roughly seven months, dated to 14,223 ± 58 years before present, was interred with two adult humans and grave goods (Janssens et al., 2018). Genome sequencing of 27 ancient dogs indicates that all dogs share common ancestry from a single now-extinct wolf population, and that genetically distinct dogs were present in Europe by about 14,200 years ago (Bergström et al., 2020). Comparison of ancient human and dog lineages across Siberia, Beringia and North America places domestication earlier still, in Siberia by approximately 23,000 years ago, possibly while people and wolves were confined together during the Last Glacial Maximum (Perri et al., 2021). Older candidate specimens exist, including a dog-like canid from the Altai Mountains dated to about 33,000 years ago, but that individual represents a lineage that did not give rise to later dogs, and its status remains contested (Ovodov et al., 2011). What is not in dispute is that dogs were our companions long before the invention of farming.
For most of that history our understanding of dogs was practical, traditional, and often contradictory. Methods were passed down through experience and folklore, and the influential training manuals of the early twentieth century rested on dominance, correction, and physical control rather than on experimental evidence (Most, 1910/2001). While some of those methods produced compliance, many also caused stress, fear or damage to the human-dog relationship. Systematic science began to replace this inheritance only with the experimental work of Pavlov (1927) and Skinner (1938), who provided the quantitative foundation for classical and operant conditioning, and the developmental research that followed (Scott & Fuller, 1965).
This analysis evaluates the scientific evolution of canine cognition and welfare to synthesize current empirical consensus. The historical trajectory begins with Darwin’s (1872) foundational observations on canine emotion, progresses through the discovery of classical and operant conditioning (Pavlov, 1927; Skinner, 1938), and reviews the mid-twentieth-century developmental benchmarks established by Scott and Fuller (1965). The framework then examines the late-1990s expansion into cognitive mechanisms (Hare et al., 2002; Miklósi et al., 1998; Topál et al., 1998) before analyzing contemporary welfare science that directly measures the impact of training methodologies (Casey et al., 2021; Vieira de Castro et al., 2020; Ziv, 2017). Finally, Section 12 explicitly outlines the limitations of the current evidence; Section 13 critically evaluates the primary arguments supporting aversive training methods; Section 14 integrates these findings across scientific, ethical, and practical knowledge translation domains; and, Section 15 argues convincingly that reward-based training is the best path forward for both results and welfare.
The scientific understanding of canine behaviour evolved from broad evolutionary observations into precise, measurable principles of learning. By shifting the focus from subjective interpretation to objective analysis, early researchers laid the groundwork for modern animal psychology and behaviour modification.
Charles Darwin treated dogs as important subjects for understanding emotion, instinct and the continuity between humans and other animals. In The Expression of the Emotions in Man and Animals, Darwin (1872) described canine facial expressions, body postures, and vocalisations in detail, arguing that many emotional expressions are shared across species. That observational tradition placed dogs at the centre of early comparative psychology, but it lacked controlled experiment. The decisive breakthrough came from physiology.
Ivan Pavlov (1849 to 1936), a Russian physiologist, was studying digestion in dogs when he observed that they salivated not only when food was placed in their mouths but in response to stimuli that reliably preceded feeding. He showed that a neutral stimulus paired repeatedly with food comes itself to elicit salivation, a process now known as classical or Pavlovian conditioning (Pavlov, 1927). Pavlov received the 1904 Nobel Prize in Physiology or Medicine for the digestive research from which this work arose.
The finding changed how learning was understood. Animals form associations between events in their environment, and those associations alter physiological and behavioural responses (Pavlov, 1927). The dogs in Pavlov's laboratory were not stubborn or wilful; they were learning the predictive structure of their world. Classical conditioning remains one of the two foundational pillars of modern learning theory, the other being operant conditioning (Skinner, 1938). Training methods built on force and correction ignored both mechanisms, treating unwanted behaviour as something to suppress rather than as information about what the dog had learned (Most, 1910/2001).
Behaviourism emphasised observable behaviour and the effects of consequences. Skinner's analysis of operant conditioning, setting out how reinforcement and punishment shape behaviour, provided the second major framework for understanding learning (Skinner, 1938). Although Skinner worked primarily with rats and pigeons, the principles apply across species, including dogs.
Dogs featured centrally in research that identified learned helplessness. In seminal experiments at the University of Pennsylvania, Overmier and Seligman (1967) and Seligman and Maier (1967) demonstrated that dogs exposed to inescapable electric shocks in a harness subsequently failed to escape shocks in a shuttle box, even when an easy escape route was freely available. Conversely, dogs that were initially trained to terminate the shock acquired the escape response normally. Exposure to inescapable shock produced this interference across a range of conditions, and the effect could not be explained by adaptation to shock or by the acquisition of competing responses (Overmier & Seligman, 1967). The operative variable was the uncontrollability of the aversive event rather than its intensity, a finding of direct relevance to any training method that relies on pain or intimidation.
Throughout this period, practical training, especially for working and military dogs, remained rooted in compulsion. Written in Germany by Colonel Konrad Most (1910/2001) and later translated into English, this foundational military-centric manual heavily emphasized dominance, physical correction, and strict control. Notably, it omitted empirical data tracking, objective statistical methodologies, and experimental control group percentages. Rather than relying on scientific validation, the text outlined a programmatic, step-by-step methodology centered on "compulsory inducements" such as the application of spiked collars, whips, and direct physical force to secure behavioural compliance from canine subjects. Consequently, Most's framework dictated the trajectory of European and North American dog training for the subsequent half-century. Science and practical training were still largely separate worlds.
One of the most influential research programmes on dogs took place at the Jackson Laboratory in Bar Harbor, Maine, under John Paul Scott and John L. Fuller. This was a massive, landmark 13-year genetic and developmental study on 250 purebred dogs. Full of rigorous statistical data, it focused heavily on genetics, social development, and emotional timelines of puppies and culminated in Genetics and the Social Behavior of the Dog (Scott & Fuller, 1965).
Among their most important contributions was the identification of sensitive, or critical, periods in puppy development. They described stages of socialisation during which puppies are especially open to forming attachments and learning about the social and physical world, and showed that experience during those windows, or its absence, has lasting effects on adult behaviour, fearfulness and social competence (Scott & Fuller, 1965). The work also examined breed differences and the interaction of genes with environment. It remains the foundation for modern understanding of puppy development and socialisation, and modern research on early adversity continues to confirm its central claim that early experience shapes later emotional functioning (Buttner et al., 2023).
Beginning in the late 1990s, empirical interest in canine cognition experienced an unprecedented global expansion, driven primarily by the pioneering frameworks of two landmark research institutions.
Founded in 1994 at Eötvös Loránd University in Budapest, the Family Dog Project revolutionized canine behavioural science. Early experiments by the group provided foundational proof that dogs intuitively rely on human communicative cues to find hidden food (Miklósi et al., 1998) and pioneered the application of infant attachment methodology to the human-dog bond (Topál et al., 1998). Recognizing that dogs are active, socially attuned interpreters of human signals rather than passive subordinates provides the empirical rationale for modern, reward-based training, which leverages this natural communicative capacity to foster cooperation and learning through positive reinforcement rather than physical force. Building upon this foundation, the Duke Canine Cognition Center, established in 2009, expanded this comparative research paradigm begun in earlier studies of dogs, wolves, and great apes (Hare et al., 2002; Hare & Tomasello, 2005).
Principal findings include the following:
• Communicative cues: Dogs follow human pointing and gaze to locate hidden food, outperforming great apes on comparable tasks. In original comparisons, they also outperform human-reared wolves (Hare et al., 2002; Hare & Tomasello, 2005; Miklósi et al., 1998). A later study comparing 44 dog puppies with 37 wolf puppies aged 5 to 18 weeks found that dog puppies were more attracted to humans, read human gestures more skilfully and made more eye contact with humans (Salomons et al., 2021). A published reanalysis of that dataset has argued that the difference narrows once rearing and environmental factors are more tightly controlled, so the magnitude of the domestication effect remains under active scientific discussion (Hansen Wheat et al., 2023).
• Attentional state: In a controlled series of trials in which dogs were forbidden to take visible food, dogs behaved differently according to whether the person was watching, had left the room, had turned away, was distracted or had closed her eyes. When watched, dogs took less food, approached it less directly and sat rather than lay down more often (Call et al., 2003).
• Word learning: The Border Collie Rico knew the labels of more than two hundred items and inferred the names of novel items by exclusion, retrieving them correctly four weeks later (Kaminski et al., 2004). Chaser, a Border Collie trained four to five hours daily over three years, learned the names of more than 1,000 objects (Pilley & Reid, 2011). Such dogs are rare; they are now studied as a model of exceptional performance rather than as typical of the species (Fugazza et al., 2021).
• Episodic-like memory: Dogs trained to imitate a human action on cue, then retrained to lie down instead, were still able to reproduce the action when unexpectedly asked to do so, with a decay curve characteristic of episodic memory (Fugazza et al., 2016).
• Emotional perception: Dogs presented with human or canine faces paired with vocalisations looked significantly longer at the face whose expression matched the valence of the sound, for both species, indicating cross-modal representation of emotional expression (Albuquerque et al., 2016). Comparative functional imaging in dogs and humans has identified voice-sensitive regions in the dog brain that respond differentially to emotionally valanced vocalisations (Andics et al., 2014).
These findings displaced older views that treated dogs as simple stimulus-response machines. Dogs are active, socially sophisticated learners who read human behaviour with considerable skill, and training methods that ignore this complexity are incomplete at best.
A significant methodological advance came from the Dog Project at Emory University. In a breakthrough starting in 2011 and 2012, researchers successfully used positive reinforcement to train pet dogs to enter and remain completely still, unrestrained, and fully conscious inside an MRI scanner (Berns et al., 2012). This produced the first functional MRI images of fully awake, cooperative dogs, revealing that the caudate nucleus, a key brain region tied to reward, activates differentially in response to human hand signals predicting food versus those signaling no reward.
Later studies from the same group found that the scent of a familiar human produced stronger caudate activation than the scent of a familiar dog or an unfamiliar person (Berns et al., 2015). In 15 dogs, the ventral caudate responded equally or more strongly to a cue predicting praise than to one predicting food in most animals tested, and those neural preferences corresponded to the dogs' behavioural choices between their guardian and a food bowl (Cook et al., 2016). When applied to a working context, researchers evaluated 49 dogs entering assistance training and found that neural activation patterns in the caudate nucleus and amygdala could accurately predict training success (Berns et al., 2017). Notably, when dogs were exposed to behavioural signals given by a stranger, their brain responses fueled a predictive classifier that achieved a 94% positive predictive value for passing the program. Beyond the data, this line of research established a profound practical paradigm: these complex neuroimaging studies are possible only because the dogs cooperate voluntarily, a level of cooperation achieved entirely through positive, low-stress training methods (Berns et al., 2012).
This voluntary, low-stress cooperation stands in stark contrast to traditional methods; when aversive training tools, such as shock collars, prong collars, or physical corrections are introduced, they trigger acute physiological and psychological stress reactions that directly undermine this cooperative bond (Schilder & van der Borg, 2004). By intentionally eliciting fear and pain, aversive techniques elevate cortisol and adrenaline levels, activating the amygdala's defensive survival circuits and conditioning the dog to view the handler as a source of distress rather than safety (Vieira de Castro et al., 2020). For policymakers and canine professionals, this neurological and physiological shift highlights a critical welfare concern demonstrated across multiple literature evaluations: physical punishment does not just suppress unwanted behaviour temporarily; it actively jeopardizes the physical and mental health of the animal while eroding the trust required for true learning (Ziv, 2017).
Dogs descend from a now-extinct population of Pleistocene wolves, and at least five genetically distinct dog lineages existed by 10,000 to 11,000 years ago (Bergström et al., 2020). Domestication involved selection for reduced fear of humans and for the social skills that support cooperation with them (Perri et al., 2021; Salomons et al., 2021). Comparisons of dog and wolf puppies raised under similar conditions report that dog puppies use human communicative gestures more readily from an early age, which points to genetic change during domestication rather than to lifetime experience alone (Hare et al., 2002; Salomons et al., 2021), though as noted in section 5 the magnitude of that difference is disputed (Hansen Wheat et al., 2023).
Breed differences in behaviour are real but routinely overstated. In the largest study of the question, surveying guardians of 18,385 purebred and mixed-breed dogs and genotyping 2,155, most behavioural traits were heritable, yet breed explained only about nine per cent of behavioural variation between individual dogs. Behaviours perceived as characteristic of modern breeds appear to derive from polygenic adaptation long predating breed formation, with modern breeds distinguished principally by appearance (Morrill et al., 2022). Practical implication: breed is a weaker predictor of how an individual dog will behave, and training expectations should be set accordingly.
Early adversity has measurable and lasting effects. Dogs rescued from adverse early environments, tested alongside a comparison group in a social stress-buffering paradigm involving a threatening stranger, showed different patterns of contact-seeking, social referencing and salivary cortisol change depending on whether their guardian was present (Buttner et al., 2023). Early experience fundamentally shapes how a dog perceives and utilizes a human as a source of security when facing a threat. While the underlying molecular and epigenetic mechanisms driving these long-term behavioural changes have been characterized primarily in rodent and human models rather than directly in canines, this mechanism is reasonably inferred by biological analogy. Crucially, the practical implication for canine welfare remains unchanged: gentle, positive early handling and comprehensive socialization are vital foundational needs, whereas early exposure to fear and pain are not neutral events but lasting developmental hazards (Buttner et al., 2023; Scott & Fuller, 1965).
Modern training is grounded in two complementary learning processes: classical conditioning, as described by Pavlov (1927), and operant conditioning, as described by Skinner (1938).
Classical conditioning builds associations between stimuli. When a clicker or verbal marker reliably predicts a reward, it becomes a conditioned reinforcer. So that a once-neutral cue gains signal value as the dog learns the predictive relationship (Pavlov, 1927).
Operant conditioning concerns the influence of consequences on behaviour (Skinner, 1938). Behaviours followed by desirable outcomes, that is, positive reinforcement, become more likely. Behaviours followed by aversive outcomes, that is, positive punishment, or by the removal of something desirable, that is, negative punishment, become less likely. Negative reinforcement, the removal of an aversive stimulus to increase behaviour, is also possible, but it carries welfare risk because the learner must first experience discomfort, fear, or pain for its removal to function as reinforcement (Ziv, 2017).
Reward-based training leverages food, play, praise, and access to valued resources to systematically build desired behaviours. Far from mere bribery, this approach relies on fundamental learning mechanics: once a behaviour is reliably established, the reinforcement schedule is gradually thinned, placing the action under precise stimulus control (Skinner, 1938). This methodology aligns with, rather than opposes, the dog’s natural social motivation, an intrinsic drive that is measurable at the neural level (Cook et al., 2016).
The two forms of punishment must be distinguished. Positive punishment adds something the dog finds unpleasant, such as a leash correction, a harshly delivered reprimand, a physical correction, or an electric stimulus, to decrease behaviour (Skinner, 1938; Ziv, 2017). Negative punishment removes something the dog wants, such as attention or access to a toy, for the same purpose. Traditional and "balanced" training methodologies rely on positive punishment and negative reinforcement to modify behaviour. Reflecting the widespread nature of these practices, one comprehensive survey revealed that 72% of respondents reported using some form of positive punishment, manifested through specific aversive equipment and physical methods such as leash jerks, vocal reprimands, and specialized deterrent hardware (Blackwell et al., 2008). Researchers determined that although owners utilizing aversive or punishment-based techniques reported behavioural compliance, these dogs exhibited severely damaged confidence and generalized anxiety. Additionally, the dogs had significantly lower proficiency in learning novel tasks and demonstrated a higher prevalence of undesirable behaviours such as, snapping, growling, or resource guarding against the owner; lashing out, barking, or lunging at unfamiliar dogs or strangers outside the household; attempting to escape when confronted with novel sounds, objects, or people; shutting down or offering zero behaviour during training out of fear of making a mistake; and, quickly abandoning the training task, failing to offer focus, and refusing to engage with the owner or the training apparatus compared to conspecifics trained via reward-based methods (Blackwell et al., 2008; Rooney & Cowan, 2011).
Epidemiological survey data reveals that the employment of physical punishment or dominance-based interventions specifically confrontational maneuvers such as "alpha rolls" or physical pinning is directly associated with a two-fold to four-fold increase in the probability of a canine displaying defensive aggression, snapping, or biting directed at the owner (Casey et al., 2013, 2014; Herron et al., 2009). Investigating the acute physiological stress response in companion dogs, Vieira de Castro et al. (2020) determined that physically invasive training methods such as leash jerking and manual bodily corrections triggered profound spikes in post-training salivary cortisol. Conversely, subjects trained using positive reinforcement maintained stable baseline stress markers, indicating that reward-based approaches preserve short-term animal welfare.
Modern force-free approaches avoid positive punishment while sometimes using carefully applied negative punishment, such as briefly pausing a game, which does not involve the addition of pain, fear, or physical force (American Veterinary Society of Animal Behavior [AVSAB], 2021).
Although aversive methods can successfully suppress behaviour in the short term, allowing them to function for generations in the limited sense that dogs stop an action to avoid immediate unpleasantness, this narrow outcome is incomplete because such techniques jeopardize the physical and mental well-being of dogs (Ziv, 2017). The critical scientific and ethical question facing professionals and policymakers is not whether punishment can interrupt behaviour, but at what severe cost to the dog's emotional state, cognitive learning capacity, trust in the handler, and long-term welfare (Schilder & van der Borg, 2004).
In the next section, this question is addressed by evaluating a robust body of empirical literature spanning stress physiology (Haverbeke et al., 2008; Schilder & van der Borg, 2004; Vieira de Castro et al., 2020), canine mood states (Casey et al., 2021), and reported behaviour problems (Blackwell et al., 2008; Herron et al., 2009; Hiby et al., 2004). Furthermore, it examines the direct impacts of training methodology on the quality of the human-dog relationship (Deldalle & Gaunet, 2014; Vieira de Castro et al., 2019) alongside evidence from controlled efficacy trials (China et al., 2020; Cooper et al., 2014).
The strength of this literature comes from convergence. Across multiple independent studies including survey research, direct behavioural observation, salivary cortisol measures, cognitive bias testing, and controlled field trials, research groups in different countries using different methods reach the same conclusion (Ziv, 2017).
In the largest quasi-experimental study of companion dog training to date, 92 dogs were recruited from three reward-based and four aversive-based training schools in Portugal. Dogs from the high-aversive group displayed more stress-related behaviours, were more frequently in tense and low behavioural states, panted more during training, and showed greater post-training increases in salivary cortisol than dogs from the reward-based group. Dogs from the mixed schools, which used aversives in lower proportions, also displayed more stress-related behaviours and more tense states than the reward-based group (Vieira de Castro et al., 2020).
Observational work on working dogs shows the same pattern. In guard-dog training of German Shepherds, direct reactions of 32 dogs to 107 shocks included lowered body posture, high-pitched yelps, avoidance, redirected aggression and tongue flicking; dogs with a history of shocks showed lower ear position and more lip licking than controls even in sessions where no shocks were delivered (Schilder & van der Borg, 2004). In an evaluation of 33 Belgian military dog handler teams, dogs adopted lower postures following aversive stimuli, and teams whose handlers used aversive stimuli less frequently performed better (Haverbeke et al., 2008). Investigating the direct behavioural fallout of operant conditioning frameworks, Deldalle and Gaunet (2014) conducted a comparative analysis of companion dogs across two training academies. The authors determined that dogs trained predominantly through negative reinforcement were more than eight times more likely to manifest acute physiological signs of distress (65%) than those trained via positive reinforcement (8%). Deldalle and Gaunet (2014) also noted striking disparities during specific cues: while reward-based schools fostered close eye contact and high attentiveness, aversive-school dogs completely ceased handler-gaze behaviours during exercises like walking on-leash, showcasing that pressure-and-release mechanics trigger avoidance-adjacent psychological states rather than mutual social engagement.
The negative welfare impact of aversive training extends far beyond the active training environment. In a landmark Portuguese study, investigating whether localized training frameworks alter a canine’s overarching emotional disposition, Vieira de Castro et al. (2020) subjected 92 companion dogs to short-term welfare tracking and subsequently moved a sub-cohort of 79 subjects to an unfamiliar location to perform a cognitive bias test. Vieira de Castro et al. (2020) demonstrated that dogs trained primarily with positive punishment and negative reinforcement were statistically more likely to interpret ambiguous environmental cues negatively. Specifically, their prolonged response latency during the spatial tasks indicated a deep-seated pessimistic bias, providing empirical evidence that high-aversive methods compromise welfare far beyond the original training context.
Using a matched-pair experimental design evaluating 100 domestic dogs in the United Kingdom, researchers demonstrated that animals whose handlers reported the implementation of two or more aversive training techniques exhibited significantly prolonged approach latencies toward ambiguous spatial locations during a standardized judgment bias paradigm compared to counterparts trained exclusively via reward-based methods. This behavioural hesitation was statistically significant at both the center target point (Wilcoxon Z = -2.38,p = .017) and the near-positive target point (Wilcoxon Z = -2.45, p = .014), providing clear empirical evidence that chronic exposure to coercive methodologies induces a more pessimistic cognitive bias and a fundamentally degraded baseline mood state (Casey et al., 2021).
Guardian-report studies consistently associate punishment-based methods with higher reported incidence of problem behaviour. In a survey of 364 households, obedience scores correlated positively with the use of rewards but not with the use of punishment, and the number of reported problem behaviours correlated with the use of punishment (Hiby et al., 2004). A larger survey linked punishment-based methods to a higher occurrence of undesirable behaviours (Blackwell et al., 2008). Among dogs referred to a veterinary behaviour service, confrontational techniques frequently elicited aggressive responses (Herron et al., 2009). Investigating the risk factors associated with human-directed canine aggression, Casey et al. (2014) and Herron et al. (2009) demonstrated that confrontational training strategies significantly compromise owner safety. Their findings indicated that utilizing forceful tactics designed to establish physical dominance escalates the likelihood of a dog retaliating via snapping or biting by a factor of two to four, suggesting that aversive interactions trigger defensive behavioural reactions rather than voluntary compliance. Higher reported use of punishment has been associated with less social interaction with strangers and reduced playfulness, while higher use of rewards predicted better performance on a novel training task (Rooney & Cowan, 2011).
Empirical evidence indicates that domestic canine subjects exposed to frequent applications of positive punishment or negative reinforcement exhibit significantly elevated frequencies of attention-seeking behaviours (Blackwell et al., 2008). Within behavioural literature, these heightened demands for contact are hypothesized to stem from the development of an insecure attachment style or a chronic, compensatory need to hyper-vigilantly monitor the handler to anticipate and avoid impending physical or verbal corrections (Vieira de Castro et al., 2019). Attachment security tends to be more consistent in dogs trained with reward-based methods (Vieira de Castro et al., 2019).
The common industry claim that aversive tools are necessary to achieve behavioural reliability does not withstand controlled empirical testing. In a rigorous study evaluating 63 dogs with established off-leash issues like poor recall, dogs were divided equally into three distinct groups: manufacturer-nominated trainers using electronic collars, those same trainers utilizing reward-based techniques without electronic stimuli, and independent trainers focused entirely on positive reinforcement (China et al., 2020). Within this design, each dog received up to 150 minutes of instruction over a five-day period. The researchers found that dogs in the positive reinforcement group responded to "sit" and "come" cues with significantly greater reliability after just a single command, alongside shorter response latencies. Consequently, China et al. concluded that electronic collar training is entirely unnecessary to achieve reliable behavioural compliance. An earlier controlled comparison conducted by the same research group reached matching conclusions, demonstrating that aversive methods compromise canine welfare without providing any advantage in training efficacy (Cooper et al., 2014).
A systematic review of 17 observational and experimental studies, Ziv (2017) concluded that aversive training methodologies significantly jeopardize the physical and psychological welfare of domestic dogs, yielding no evidence that positive punishment exceeds the efficacy of reward-based training. Specifically, the data showed that 25% to 43% of dogs subjected to dominance-based manipulations retaliated with human-directed aggression, whereas reward-trained subjects maintained significantly higher task-learning scores and lower behavioural regression rates. Additionally, the review identified multiple cohorts demonstrating superior outcomes for positive reinforcement; for example, owners utilizing reward-based training reported a 10% to 20% increase in obedience scores and a concurrent 30% reduction in subsequent behavioural problems compared to cohorts utilizing physical corrections (Ziv, 2017). Research comparing reward and mixed methods on efficiency as well as efficacy reached comparable conclusions (Vieira de Castro et al., 2021).
While force-based methods may have produced immediate compliance in the past, extensive modern behavioural research demonstrates that they compromise canine welfare across three primary domains. Physically invasive tactics induce acute physiological and behavioural stress, characterized by significant elevations in post-training salivary cortisol levels and immediate postural collapse (Deldalle & Gaunet, 2014; Vieira de Castro et al., 2020). Furthermore, these techniques elicit profound fear, frequently manifesting as defensive aggression, a pessimistic cognitive bias, or an escalation in bite risk toward the handler (Casey et al., 2021; Herron et al., 2009). Systemic utilization of positive punishment results in long-term relationship damage, characterized by degraded gaze-attentiveness, an insecure owner-dog attachment bond, and a generalized avoidance of the handler (Deldalle & Gaunet, 2014; Vieira de Castro et al., 2019; Ziv, 2017).
· The Positions of Behavioral Specialists: Veterinary societies dedicated specifically to animal behaviour and clinical ethology maintain the strictest positions against aversives. The American Veterinary Society of Animal Behavior (AVSAB, 2021) issued an authoritative position statement asserting that only reward-based methods should be utilized for training everyday skills and modifying challenging behaviours. AVSAB (2021) noted that there is no scientific evidence justifying the necessity of aversive tools, a stance they reaffirmed to clarify that aversive punishment should not be used under any circumstances (AVSAB, 2025). Similarly, the American College of Veterinary Behaviorists (ACVB, n.d.) stands firmly against any methodology that induces short- or long-term pain, discomfort, or fear. The ACVB (n.d.) warns that such tools pose a profound threat to animal welfare by actively inhibiting learning, causing direct physical injury, and triggering behavioural escalations related to fear and distress. The European Society of Veterinary Clinical Ethology (ESVCE) echoes these principles, advocating for non-aversive behaviour modification rooted in humane clinical ethology and scientific learning theory.
· Hospital and Practice-Specific Guidelines: General and companion animal veterinary practice associations have formalized these behavioural stances into clinical operational standards. The American Animal Hospital Association (AAHA), via its Canine and Feline Behavior Management Guidelines, actively opposes training techniques that rely on physical or psychological confrontation (AAHA, 2015). AAHA explicitly details that aversive methods cause severe disruptions to the human-animal bond, damage problem-solving capabilities, and can accelerate behavioural disorders in already distressed patients (AAHA, 2015). In the United Kingdom, the British Small Animal Veterinary Association (BSAVA, 2022) mirrors this directive. The BSAVA (2022) officially recommends against tools meant to deliver an unpleasant stimulus such as electric shock collars, prong collars, spray collars, choke chains, and physical corrections like lead jerking citing that they unacceptably compromise canine welfare.
· Transnational and Global Coalitions: Recognizing the global nature of animal welfare concerns, several large networks have consolidated their positions to provide unified policy updates. A landmark initiative was achieved when the Federation of Veterinarians of Europe (FVE), the Federation of European Companion Animal Veterinary Associations (FECAVA), the Federation of European Equine Veterinary Associations (FEEVA), and the World Small Animal Veterinary Association (WSAVA) co-authored a joint position paper regarding animal training (FVE et al., 2024). This international coalition outlines 14 evidence-based recommendations designed to optimize welfare. Their joint stance heavily mandates positive reinforcement-based methods while structurally opposing aversive infrastructure, singling out tools like electric pulse collars as unacceptable in modern animal care (FVE et al., 2024).
· National Veterinary Associations: Domestic veterinary governing bodies have aligned their national policies with this global consensus. The Canadian Veterinary Medical Association (CVMA, 2021) officially declared that aversive methods are strongly discouraged because they fail to address the underlying psychological drivers of an unwanted behaviour while inducing fear, distress, anxiety, or physical pain. To combat this, the CVMA (2021, 2024) actively supports establishing national curriculum benchmarks for non-aversive trainers, aligning directly with professional codes of ethics that place severe, outdated training equipment on restricted lists. Furthermore, the British Veterinary Association (BVA), the Australian Veterinary Association (AVA), and the New Zealand Veterinary Association (NZVA) maintain official position papers advocating for force-free training, jointly concluding that punitive methodologies are counterproductive to public safety and modern veterinary welfare targets.
· North American Humane Societies’ Positions: In the United States, the American Society for the Prevention of Cruelty to Animals (ASPCA) outlines a strict framework for evaluating training protocols. The ASPCA (n.d.) explicitly opposes any training equipment or methodology that causes an animal to experience physical discomfort or undue anxiety. Instead, the organization champions techniques rooted in an understanding of behavioural science, prioritizing lures and positive reinforcement such as food, play, and praise to manage and modify behaviour with the lowest possible stress threshold (ASPCA, n.d.). Similarly, the Humane Society of the United States (HSUS) warns pet owners against tools that rely on discomfort or fear, specifically naming prong collars, choke chains, and electric shock collars as equipment that damages the human-animal bond and risks escalating fear into defensive aggression.
In Canada, systemic opposition to aversives is heavily integrated into animal protection policies. The British Columbia Society for the Prevention of Cruelty to Animals (BC SPCA), which serves as a benchmark for national standards via Humane Canada, maintains a comprehensive position statement rejecting coercion and force (BC SPCA, 2016). The BC SPCA (2016) asserts that while punishment-based aversive techniques may suppress an unwanted behaviour in the short term, they fundamentally fail to treat the underlying psychological trigger. Consequently, these practices cause undue distress, anxiety, pain, and physical injury (BC SPCA, 2016). To actively combat these issues, the organization enforces strict equipment restrictions, explicitly opposing tools like choke chains, prong collars, shock collars, bullhooks, and physical actions such as striking or pinching (BC SPCA, 2016).
· United Kingdom and Oceana Humane Societies’ Positions: The Royal Society for the Prevention of Cruelty to Animals (RSPCA) across the United Kingdom and Australia actively campaigns against the physical and psychological toll of punitive training. In the United Kingdom, the RSPCA focuses heavily on the ethical implications of tools meant to inflict pain or discomfort. The RSPCA UK (n.d.-b) explicitly states that aversive training techniques compromise long-term mental well-being and overall happiness. Furthermore, the organization actively lobbies for legislative bans on specific tools, emphasizing that tools like prong collars cause severe puncture wounds and nerve damage, while confusing the animal rather than teaching alternative, desirable behaviours (RSPCA UK, n.d.-b). The RSPCA UK (n.d.-a) consistently advocates for legislative measures to eliminate electric shock collars entirely, branding them as cruel and entirely unnecessary alternatives to positive reinforcement.
Consistently, RSPCA Australia codifies these restrictions into formal animal welfare policies. The organization defines aversive equipment including shock-delivery mechanisms, anti-barking devices, invisible boundaries, prong collars, and choke chains as inherently inhumane (RSPCA Australia, 2023, 2025). According to official guidelines, such tools rely heavily on fear, intimidation, and punishment, which significantly increases the risk of injury to both the animal and the handler (RSPCA Australia, 2025). Furthermore, these policies extend beyond companion animals, explicitly opposing any electrical shock devices or aversive equipment meant to modify behaviour or suppress stereotypic habits in performance, sport, entertainment, or working animals (RSPCA Australia, 2026).
In New Zealand, the state-recognized animal welfare authority maintains strict, legally backed parameters regarding behavioural modification. The Royal New Zealand Society for the Prevention of Cruelty to Animals (SPCA New Zealand, 2024) enforces an explicit position statement mandating low-stress, force-free training utilizing reward-based learning. The organization formally opposes any method or equipment designed to apply an unpleasant or painful stimulus to an animal to suppress behaviour (SPCA New Zealand, 2024). Specifically, SPCA New Zealand (2024) target a broad spectrum of everyday tools, designating electric shock collars, ultrasonic devices, citronella spray aids, acoustic noise makers, spray bottles, prong collars, and choke chains as unacceptable. Furthermore, the organization actively lobbies the New Zealand government to institute total national prohibitions on the sale, importation, and use of all electronic shock-delivery mechanisms.
· African Humane Societies’ Positions: In South Africa, statutory animal protection networks utilize criminal statutes to eliminate punitive methodologies. The National Council of SPCAs (NSPCA) coordinates with domestic sport and kennel authorities to enforce strict codes of conduct. Through joint structural enforcement with bodies like the Kennel Union of South Africa (KUSA, 2024), the use of prong, spiked, or electric shock collars is explicitly classified as a severe breach of animal protection regulations. Under these binding national notices, all affiliated training facilities, clubs, and behaviourists are prohibited from utilizing physical corrections, with the organization mandating immediate disciplinary and legal intervention against individuals employing aversive infrastructure (KUSA, 2024).
· European and Transnational Animal Protection Societies Organisations’ Positions: In Ireland, domestic welfare policies are tightly integrated with broader European force-free initiatives. The Irish Society for the Prevention of Cruelty to Animals (ISPCA, n.d.) standardizes its rescue, rehabilitation, and rehoming procedures entirely around positive reinforcement husbandry. The ISPCA (n.d.) explicitly discourages the consumer use of remote or automated aversive tools, aligning its legislative lobbying efforts with European coalitions to classify pain-inducing training equipment as restricted or prohibited devices.
On a continental scale, these individual national stances are consolidated by Eurogroup for Animals, a powerful federation representing over 100 animal protection organizations across Europe. Eurogroup for Animals (2024, 2025) structurally lobbies the European Parliament and Member States to embed strict, science-based behaviour and handling guidelines into overarching European Union animal welfare legislation. The coalition actively supports the complete phase-out and legal prohibition of aversive equipment citing successful legislative bans in nations such as Austria, Germany, Denmark, Sweden, and France while systematically endorsing the institutionalization of positive reinforcement and force-free handling protocols across all European member organizations (Eurogroup for Animals, 2024, 2025).
Where such statements are relied upon in a legislative submission, each should be cited individually to its own current published version, since wording and dates are revised periodically.
Mixed or balanced training does not escape these findings. Comparisons of high-aversive, mixed, and reward-only groups show that even lower proportions of aversive technique are associated with elevated stress indicators relative to reward-based training alone (Vieira de Castro et al., 2020).
For most of recorded history animals were treated primarily as property. Scientific and philosophical views changed gradually. Darwin's comparative work on emotional expression (Darwin, 1872), the growth of ethology (the scientific study of animal behaviour), and later findings in neuroscience and cognitive science strengthened the case that mammals, including dogs, are sentient, that is capable of feeling and of having experiences that matter to them. Ethologist and Behavioural Ecologist Marc Bekoff has argued on the basis of behavioural observation and comparative evidence that many animals, dogs among them, experience a range of emotions including joy, fear, grief and attachment, and that recognising these states is central to understanding them as sentient beings (Bekoff, 2007).
Experimental evidence now supports that position directly. Dogs form attachment bonds with human caregivers that functionally resemble infant and caregiver attachment, using the caregiver as a secure base and safe haven (Topál et al., 1998). Neuroimaging has mapped reward and emotional responses in the canine brain, including differential activation to caregiver-related odours and to praise (Berns et al., 2015; Cook et al., 2016). Comparative imaging has identified voice-sensitive regions responsive to emotional vocalisations (Andics et al., 2014). Dogs integrate facial and vocal cues to categorise emotional expressions across species (Albuquerque et al., 2016), and cognitive bias testing shows that affective state influences their decision-making (Casey et al., 2021; Vieira de Castro et al., 2020). Taken together, this work moves the question of dog sentience from philosophical speculation to empirical science.
Legal recognition followed more slowly. A decisive turning point occurred at the European level when the Protocol on Protection and Welfare of Animals, annexed to the Treaty of Amsterdam in 1997, legally recognized animals as sentient beings for the first time (Treaty of Amsterdam, 1997). This landmark agreement mandated that the European Community and its member states fully incorporate animal welfare requirements when shaping public policy. This foundational recognition was subsequently integrated into the core of European law as Article 13 of the Treaty on the Functioning of the European Union (taking effect in December 2009), which significantly broadened the scope of protected policy fields (Treaty on Kisan/Functioning of the European Union, 2007/2009, Art. 13).
Building directly upon this legal foundation of animal sentience, Article 13 of the Treaty on the Functioning of the European Union (2012) has been translated by at least 13 European nations into explicit legislative prohibitions against aversive training tools. Recognizing that electronic shock collars, prong collars, and physical deterrents cause unnecessary pain and psychological distress incompatible with sentient welfare, countries such as Denmark, Norway, Sweden, Finland, and Iceland (Nordic Block); Germany, France, Austria, Switzerland, and the Netherlands (Western Europe); Spain and Slovenia (Southern and Eastern Europe); and, Wales (United Kingdom) have enacted comprehensive national bans or severe restrictions on their sale and use. For policymakers, this trajectory illustrates a clear regulatory shift: when the law formally acknowledges animals as sentient, the continued public authorization of pain-based training equipment becomes legally and ethically untenable.
This constitutional and statutory trend is exemplified in Germany, which elevated animal protection to a constitutional state objective under Article 20a of its Basic Law by explicitly inserting the phrase "and the animals" in 2002 (Grundgesetz [Basic Law], Art. 20a). Similarly, the United Kingdom codified animal sentience through the Animal Welfare (Sentience) Act 2022, establishing an Animal Sentience Committee dedicated to evaluating and reporting on how government policies impact animals as conscious, feeling beings (Animal Welfare (Sentience) Act 2022, c. 22). As other jurisdictions introduce comparable statutory or constitutional frameworks, each policy measure must be mapped directly to its respective governing legal instrument to maintain regulatory accountability and alignment with modern animal welfare science.
This change has direct consequences for training and management. Once animals are acknowledged as sentient, methods that deliberately inflict pain, fear or distress require stronger justification. The same ethical logic that supports restrictions on certain farming practices supports restrictions on electric shock collars, prong collars, and comparable devices. Jurisdictions that have restricted these tools are acting consistently with the recognition of sentience: if a dog can suffer, causing unnecessary suffering in the name of training is difficult to defend when effective alternatives with lower welfare cost exist (American Veterinary Society of Animal Behavior [AVSAB], 2021; China et al., 2020).
Recognition of sentience does not by itself dictate the detail of training practice. It raises the standard of justification required for any method that risks causing fear, distress, or pain. The scientific evidence summarised in this position paper supplies the practical grounding for that standard.
Dogs are highly social, cognitively sophisticated animals whose minds have been shaped by domestication to attend to and communicate with humans (Hare et al., 2002; Hare & Tomasello, 2005; Miklósi et al., 1998; Salomons et al., 2021). They learn through association and through the consequences of their actions (Pavlov, 1927; Skinner, 1938), retain memory of incidentally encoded events (Fugazza et al., 2016), attend to human attentional state (Call et al., 2003), and in exceptional individuals acquire extensive referential vocabularies (Fugazza et al., 2021; Kaminski et al., 2004; Pilley & Reid, 2011).
Early experience, particularly during sensitive developmental periods, has lasting effects on emotionality, social behaviour, and stress reactivity (Buttner et al., 2023; Scott & Fuller, 1965). Individual and breed differences exist but breed accounts for a modest share of behavioural variation between individuals, and plasticity and learning remain central throughout life (Morrill et al., 2022).
Dogs experience emotions including fear, anxiety and positive affective states such as joy, contentment, playfulness, curiosity, and a sense of security or comfort. Moreover, they perceive emotional expression in humans and in other dogs (Albuquerque et al., 2016; Andics et al., 2014; Bekoff, 2007). Chronic or intense aversive experience produces changes in behaviour and welfare visible in physiological markers and in cognitive bias (Casey et al., 2021; Schilder & van der Borg, 2004; Vieira de Castro et al., 2020). Dogs form attachment relationships with human caregivers that influence security, exploration, and recovery from stress (Buttner et al., 2023; Topál et al., 1998).
Training is therefore not merely the suppression of unwanted behaviour. It is the construction of a shared communication system, the building of skills, and the maintenance of a cooperative relationship. Methods that undermine trust or elevate fear work against these goals (Herron et al., 2009; Vieira de Castro et al., 2019).
Traditional methods based on positive punishment and compulsion have been used for generations (Most, 1910/2001), and they often produce the appearance of success while the handler is present, and the threat of correction remains credible. That outcome is not imaginary. The field is moving away from these methods not because they never worked, but because they work at a cost the dog pays in fear, stress, damaged trust and, in some cases, increased aggression or behavioural shutdown (Schilder & van der Borg, 2004; Herron et al., 2009; Ziv, 2017).
11.1 Effectiveness
Reward-based training builds new behaviours and addresses unwanted behaviour by teaching incompatible alternatives and altering the antecedents and consequences that maintain the problem. There is no robust evidence that adding positive punishment improves outcomes for companion dogs, and controlled comparison indicates the opposite for common problems such as recall (Cooper et al., 2014; China et al., 2020; Ziv, 2017; AVSAB, 2021).
11.2 Welfare
Reward-based methods minimise fear, pain and distress and keep the dog in a state more conducive to attention and voluntary engagement (Deldalle & Gaunet, 2014; Rooney & Cowan, 2011). Empirical evaluations tracking physiological indicators, behavioural body language, and cognitive judgment biases consistently demonstrate that the application of positive punishment and high-intensity aversive pressure inflicts severe, measurable welfare costs on canines (Casey et al., 2021; Haverbeke et al., 2008; Schilder & van der Borg, 2004; Vieira de Castro et al., 2020).
11.3 Relationship
Reward-based training becomes a source of positive interaction rather than conflict, and dogs trained this way show greater willingness to work with their handlers and more consistent attachment security (Rooney & Cowan, 2011; Vieira de Castro et al., 2019). Compulsion can teach a dog that the handler is a source of threat (Schilder & van der Borg, 2004; Herron et al., 2009), which undermines the cooperative relationship most guardians want (Deldalle & Gaunet, 2014).
11.4 Generality and Safety
These behavioural and neurological principles scale seamlessly from basic manners to advanced assistance and detection work; notably, assistance dogs conditioned under positive reinforcement protocols exhibit measurable neural response profiles that can be successfully utilized to select candidates and predict working success (Berns et al., 2017). Reward-based methods also reduce the risk of redirected aggression and shutdown associated with heavy-handed correction (Schilder & van der Borg, 2004; Herron et al., 2009).
11.5 Alignment with How Dogs Learn
Reward-based training works with classical and operant conditioning (Pavlov, 1927; Skinner, 1938), with the dog's social motivation (Cook et al., 2016), and with what is now known about canine cognition and emotion (Andics et al., 2014; Albuquerque et al., 2016). Compulsion often addresses symptoms while leaving the underlying learning and emotional processes untouched (Ziv, 2017).
Negative punishment, that is, the removal of a desired outcome, is still used thoughtfully within reward-based frameworks, for example calmly ending a game when a dog jumps up. Because it does not involve adding pain or fear, it occupies a different position on the welfare scale from positive punishment (AVSAB, 2021). Contemporary training prioritizes the removal of aversive procedures while fully leveraging reinforcement, management strategies, and environmental design. When used at all, negative punishment is applied selectively and carefully to support learning and welfare.
The validity of the conclusions presented in this position paper derives from the consistent convergence of multiple independent lines of empirical evidence rather than reliance on an isolated study. To maintain methodological transparency and academic rigor, the structural boundaries and inherent limitations of this collective data must be explicitly stated.
While not every piece of canine research relies on a controlled experiment, a robust portion of the literature derives from guardian questionnaires and observational designs; while these methodologies excel at identifying behavioural correlations, they cannot establish direct causation on their own (Blackwell et al., 2008; Herron et al., 2009; Hiby et al., 2004; Rooney & Cowan, 2011). Furthermore, definitions of aversive, reward-based, and "balanced" training frequently vary among studies and practitioners - a methodological challenge heavily documented across review literature (Ziv, 2017). This complexity is compounded by individual variations in temperament, unique behavioural histories, and underlying motivation, particularly given that breed remains a remarkably weak predictor of an individual dog's behaviour (Morrill et al., 2022). Methodologically, research lines also diverge in their primary metrics: some prioritize acute welfare indicators like stress behaviours, cortisol levels, and cognitive bias (Casey et al., 2021; Schilder & van der Borg, 2004; Vieira de Castro et al., 2020), whereas others isolate training performance and task efficiency (China et al., 2020; Vieira de Castro et al., 2021). Although welfare and performance are deeply intertwined, they are not identical metrics.
Finally, where biological mechanisms are inferred from non-canine research, such as the epigenetic impacts of early adversity detailed in the preceding section, this position paper explicitly frames those findings as cross-species analogies rather than direct canine demonstrations. The existence of effective reward-based methods does not mean that every reward-based intervention succeeds or that every practitioner applies them skilfully. Positive reinforcement does not mean indiscriminate delivery of food; it means the contingent addition of valued outcomes to strengthen desired behaviour, combined with management, environmental design, and the teaching of alternative responses (AVSAB, 2021; Skinner, 1938).
These limitations do not cancel the overall pattern. Across experimental comparisons, observational studies, physiological measures, cognitive bias tests and professional consensus, methods relying on intentional aversive stimulation are associated with higher welfare costs, while skilled reward-based approaches achieve legitimate training objectives (Casey et al., 2021; China et al., 2020; Vieira de Castro et al., 2020; Ziv, 2017). The case does not require any single study to be definitive. It rests on the consistency of direction across independent methods.
To be equitable and persuasive, a position paper must acknowledge relevant counterarguments to the opposing view and respond to them in a substantive manner. This includes arguments for “balanced” or aversive approaches. The following propositions are frequently asserted.
Critics argue that aversives are necessary for difficult or high-drive dogs. However, controlled comparison of recall and general obedience in dogs with known off-lead problems did not demonstrate that electronic collars are required to achieve the target behaviours when skilled positive reinforcement is available (China et al., 2020). The review literature reaches the same conclusion on necessity (Ziv, 2017; AVSAB, 2021). High motivation can be channelled with appropriately valued reinforcers and careful management; the claim of necessity is not substantiated by the available comparative evidence.
Opponents contend that electronic collars are more effective for recall. Nonetheless, in the controlled study described above, the reward-focused group achieved better responses to a single cue and shorter latencies than the electronic collar group on the measured tasks (China et al., 2020). Welfare measures in the earlier trial by the same group pointed the same way (Cooper et al., 2014). The data do not support a general superiority claim.
Although proponents of balanced dog training claim it delivers faster results, evidence suggests this immediate compliance stems from a state of emotional shutdown induced by acute discomfort or fear. This perspective is fundamentally flawed. Mixed training methodologies that incorporate aversive techniques significantly elevate a dog's acute and long-term stress indicators compared to purely reward-based approaches (Vieira de Castro et al., 2020). Furthermore, direct comparative research evaluates that mixed training offers no actual advantage in terms of learning efficiency or speed (Vieira de Castro et al., 2021). Speed of suppression is different from enduring, low-stress learning.
Proponents contend that aversives are harmful only when misused, asserting that correct application yields safe results. However, research indicates that the welfare cost of these techniques is not confined to extreme misuse. Although severity and contingency influence outcomes, comparative studies of training schools and guardian practices utilizing aversive methods even at low frequencies, demonstrate elevated stress-related behaviours and pessimistic cognitive biases (Casey et al., 2021; Vieira de Castro et al., 2020). Furthermore, skilled application within a professional military training context does not eliminate these adverse effects (Haverbeke et al., 2008), suggesting that technical proficiency does not neutralize the inherent risks of aversive methodology.
Beyond the argument of proper application, advocates of balanced training contend that the historical success of working, and service dogs serves as definitive proof that aversive methods are effective. Although it is undisputed that aversive techniques can produce behavioural compliance, the primary policy-relevant question is whether this potential efficacy justifies the associated welfare costs when reward-based alternatives achieve identical objectives. Accumulating literature demonstrates that specialized assistance dogs can be successfully selected and trained entirely under positive reinforcement protocols (Berns et al., 2017). Furthermore, observational research in highly demanding military environments indicates that team performance improves when handlers utilize aversive stimuli less frequently (Haverbeke et al., 2008), suggesting that force-free alternatives neutralize the operational necessity of punitive methods.
In summary, the evidence demonstrates that the core debate surrounding aversive methods does not hinge on whether punitive tools can successfully suppress behaviour. Instead, the critical policy and ethical question is whether any marginal increase in efficacy justifies the documented welfare costs, particularly when highly effective, reward-based alternatives are readily available. Research consistently confirms that the psychological toll of aversive techniques persists regardless of handler skill level or tool severity (Casey et al., 2021; Haverbeke et al., 2008; Vieira de Castro et al., 2020), while successful positive-reinforcement models in service and military sectors prove that force-free methodologies satisfy identical operational goals (Berns et al., 2017). Because reward-based protocols achieve equivalent objectives without compromising animal welfare, the continued systemic reliance on aversive training remains empirically and ethically indefensible.
From Darwin's observations (Darwin, 1872) to Pavlov's conditioned reflexes (Pavlov, 1927), from Skinner's analysis of consequences (Skinner, 1938) to the Scott and Fuller developmental timelines (Scott & Fuller, 1965), and on through modern cognitive and welfare research (Albuquerque et al., 2016; Andics et al., 2014; Berns et al., 2012; Buttner et al., 2023; Call et al., 2003; Casey et al., 2021; Fugazza et al., 2016; Hare et al., 2002; Kaminski et al., 2004; Miklósi et al., 1998; Morrill et al., 2022; Salomons et al., 2021; Topál et al., 1998; Vieira de Castro et al., 2020; Ziv, 2017), science has steadily replaced folklore with evidence. Dogs are neither blank slates nor stubborn adversaries. They are predictive learners, social partners, and emotional beings whose behaviour makes sense once their learning history, developmental needs, and the signals we send them are understood. That partnership extends back at least 14,000 years and probably much further (Bergström et al., 2020; Janssens et al., 2018; Perri et al., 2021).
Reward-based training is not a soft option. It is the approach best supported by research into how animals learn, how dogs perceive and communicate with us, and what keeps them physically and psychologically healthy (AVSAB, 2021; China et al., 2020; Ziv, 2017).
Effective animal management relies entirely on empirical evidence. By accurately reading a dog’s physiological markers including body language, facial expressions, posture, and arousal levels within a specific context, we can diagnose the true emotional and functional drivers of behaviour. A dog that growls, freezes, or fails to respond is communicating critical data. Expert observation allows handlers to identify and resolve the underlying emotional states or unmet needs driving these reactions such as fear, frustration, confusion, or deficit skills rather than merely suppressing surface symptoms through punishment (Herron et al., 2009; Ziv, 2017).
Consequently, modern training must pivot to a relationship-based framework. This methodology prioritizes positive reinforcement to construct desired behaviours, strategic environmental management to block the rehearsal of unwanted actions, and the systematic instruction of alternative skills (AVSAB, 2021). Peer-reviewed data proves that dogs trained under these humane, science-based protocols demonstrate a significantly higher willingness to work, stronger attachment security, drastically lower stress metrics, and fewer secondary behavioural complications (Hiby et al., 2004; Rooney & Cowan, 2011; Vieira de Castro et al., 2019, 2020). This approach transforms the human-canine bond into a foundational source of safety and public motivation (Cook et al., 2016; Topál et al., 1998).
Several jurisdictions have already restricted electric collars and comparable devices. Wales prohibited the use of any electronic collar designed to administer an electric shock on cats and dogs under the Animal Welfare (Electronic Collars) (Wales) Regulations 2010. Made pursuant to section 12 of the Animal Welfare Act 2006, this prohibition covers use rather than sale.
Austria’s animal welfare regulations enforce a zero-tolerance approach by banning the use, sale, and possession of aversive dog training tools. Rather than only restricting punitive methods, the law explicitly treats the use of painful or fear-inducing devices as animal cruelty (Bundesgesetz über den Schutz der Tiere [Federal Act on the Protection of Animals], 2004, § 5). Consistent with this framework, the Tierschutzgesetz (TSchG) criminalizes equipment designed to control canine behaviour through unpleasant stimuli, startle responses, or physical pain (Tierschutzgesetz [TSchG], 2004, § 5). As a result, Austria’s standard has shifted toward reward-based, force-free positive reinforcement training.
France initially addressed coercive training devices through a legislative bill (Proposition de loi n° 563) introduced by Member of Parliament Corinne Vignon. On January 16, 2023, the French National Assembly voted with an overwhelming majority (111 to 5) to pass the bill, which targeted any device delivering an electrical charge, tightening without restriction (choke chains), or utilizing internal physical spikes (prong collars). To overcome ongoing legislative gridlock within the Senate, the executive branch intervened. The French Ministry of Agriculture finalized the enforcement pipeline by publishing an administrative decree (arrêté) on June 19, 2025. This official decree formally codified strict sanitary and animal protection regulations into law, permanently prohibiting professional trainers, breeders, and shelters from utilizing or endorsing coercive, electric, or unrestricted strangle equipment.
Germany prohibits the use of devices that, by direct electrical action, restrict an animal's species-appropriate behaviour or compel it to move, thereby causing more than insignificant pain, suffering or harm (Tierschutzgesetz [Animal Welfare Act], § 3, No. 11). The Federal Administrative Court held in 2006 that this covers electronic training devices used in dog training, including modern low-current devices, and that proof of actual injury in the individual case is not required (BVerwG, 2006).
The Netherlands prohibited the use of equipment capable of delivering electric shocks to a dog by adding a new subsection to Article 1.3 of the Besluit houders van dieren (Animal Keepers Decree), in force from 1 January 2022, with narrow exceptions for veterinary procedures and defined police and military tasks. Requiring a dog to wear such a device constitutes use. Collars capable of causing pain through sharp protrusions were already prohibited under the preceding subsection (Besluit houders van dieren [Animal Keepers Decree], 2014, Art. 1.3).
The landscape of Spanish jurisprudence underwent a paradigm shift regarding companion animal equity following the enactment of Law 7/2023, of March 28, on the protection of the rights and welfare of animals (Ley de Protección de los Derechos y el Bienestar de los Animales). Designed to systematically eliminate neglect, abandonment, and coercive handling practices, Article 27 of this national statute explicitly outlaws the deployment of aggressive, behaviour-modifying hardware such as electric shock collars, electronic impulse transmitters, prong or spike collars, and physical choking mechanisms deemed capable of inflicting physical trauma or psychological distress upon an animal. Furthermore, the legislation establishes rigid thresholds for owner accountability by introducing universal mandates that require all domestic dog handlers, irrespective of the animal's breed or weight class, to successfully complete a certified training course verifying baseline comprehension of canine ethology alongside maintaining valid civil liability insurance coverage (Boletín Oficial del Estado, 2023).
Across the Nordic region, the use of electronic shock collars (e-collars) for canine behavioural modification is strictly regulated or entirely prohibited under modern animal protection frameworks. In Denmark, the statutory prohibition is executed via the Danish Animal Welfare Act (2010), which bans automatic, remote, and boundary-based electric devices. Norway takes a similar stance under the Norwegian Animal Welfare Act (2009), which categorizes electrical training tools as a source of unnecessary distress, though tightly controlled exceptions exist for livestock avoidance training. In Sweden, the comprehensive Swedish Animal Welfare Act (2018) extends its protection past electronic stimuli to also encompass physical compliance tools like prong collars. More recently, Finland enacted the sweeping Animal Welfare Act (2023), which went into effect in 2024; according to the Finnish Kennel Club (2024), this statute completely criminalizes not only the active use but also the commercial sale, import, and distribution of shock and spiked collars. Finally, Iceland maintains matching prohibitions under its updated national Animal Protection Act, completing a regional consensus that legalizes passive tracking technology, such as Global Positioning System (GPS) locators, while fundamentally outlawing aversive, shock-based training methodologies (Satellai, 2026).
Slovenia formally outlawed the use of electronic shock training collars in 2011 under its foundational animal welfare law, the Animal Protection Act (Zakon o zaščiti živali). Rather than introducing isolated, device-specific bans, Slovenian lawmakers incorporated this restriction into an expansive overhaul of their national animal protection framework.
Beyond the strict regulations in Europe, jurisdictions across Latin America and Oceania are aggressively tightening their animal welfare standards. For example, Colombia has progressive nationwide legislation that outlaws both electric shock and pronged collars. In Australia, the federal government intervenes at the border by blocking the entry of pronged devices via federal import restrictions. However, the Australian Bureau of Agricultural and Resource Economics and Sciences (ABARES) and Business Queensland document a highly fragmented internal system. Because the Australian Constitution does not centralize animal welfare, state and territory governments independently regulate training equipment. To date, state-level bans on electronic shock collars are active in several regions including Australian Capital Territory, New South Wales, Queensland, South Australia, Tasmania, and Victoria.
North America exhibits a sharp, fragmented legal contrast between its northern and southern neighbours. In Canada, subnational regulation takes priority. The International Association of Canine Professionals highlights this regional autonomy, noting that the province of Quebec officially banned electronic shock collars under its provincial animal welfare acts. Conversely, the United States completely lacks federal legislation restricting these devices. Although several state legislatures have introduced pending animal protection bills, real change has been driven by corporate market intervention rather than government mandates. A prominent example occurred when the major pet specialty retailer Petco voluntarily removed human-controlled shock collars from its commercial inventory to promote positive training models.
Because such provisions are amended periodically, current national instruments should be checked and cited directly in any submission.
In these jurisdictions, dogs continue to be trained successfully for companion life, sport, assistance work, and other roles. Continued functioning without these tools is real-world evidence that aversive equipment is not necessary, consistent with the controlled finding of no efficacy advantage (China et al., 2020) and with the review-level conclusion that aversive methods are not more effective than reward-based training (Ziv, 2017). A qualification is warranted: jurisdictional restrictions reflect policy judgements informed by welfare science. They do not themselves prove the underlying scientific claims, which rest on the empirical studies summarised in section 8.
Methods that depend on positive punishment persist in some sectors through transmission of practice (Most, 1910/2001). Continuity of practice does not establish that those methods are superior or necessary when lower-welfare-cost alternatives exist. As research continues into cognition, emotion, genetics and the neural basis of learning, the capacity to train humanely and effectively will continue to improve. The responsibility of anyone who lives or works with dogs is to keep pace with that knowledge and to choose methods that respect both the science and the animal.
The trajectory of behavioural science extending from the foundational models of evolutionary biology and classical conditioning to contemporary neurocognitive and canine welfare research has conclusively dismantled antiquated folklore in favour of empirical evidence (Albuquerque et al., 2016; Andics et al., 2014; Berns et al., 2012; Buttner et al., 2023; Call et al., 2003; Casey et al., 2021; Darwin, 1872; Fugazza et al., 2016; Hare et al., 2002; Kaminski et al., 2004; Miklósi et al., 1998; Morrill et al., 2022; Pavlov, 1927; Scott & Fuller, 1965; Salomons et al., 2021; Skinner, 1938; Topál et al., 1998; Vieira de Castro et al., 2020; Ziv, 2017). Dogs are neither blank slates nor stubborn adversaries; they are sophisticated social partners, predictive learners, and emotional beings whose co-evolutionary partnership with humans spans over 14,000 years (Bergström et al., 2020; Janssens et al., 2018; Perri et al., 2021).
Far from a passive or "soft" alternative, evidence-based, reward-based training represents the absolute pinnacle of scientific efficacy, maximizing canine learning capacity while safeguarding physical and psychological welfare (AVSAB, 2021; China et al., 2020; Ziv, 2017). By analyzing physiological markers and diagnosing the functional drivers of behaviour such as fear, confusion, or deficit skills expert handlers can resolve underlying emotional distress rather than merely suppressing superficial symptoms through physical punishment (Herron et al., 2009; Ziv, 2017). The viability of this humane paradigm is robustly verified by comprehensive real-world evidence across Wales, Austria, France, Germany, the Netherlands, Spain, Slovenia and the entire Nordic region, where sweeping statutory bans on electronic shock devices and mechanical prong collars have been enacted without compromising the successful training of companion, sport, and working assistance dogs (Animal Welfare Act, 2006, 2009, 2018, 2023; Besluit houders van dieren, 2014; BVerwG, 2006; Boletín Oficial del Estado, 2023; Danish Animal Welfare Act, 2010; Finnish Kennel Club, 2024; Ministère de l'Agriculture et de la Souveraineté alimentaire, 2025; Satellai, 2026; TSchG, 2004, § 5; Zakon o zaščiti živali, 2011).
While outdated coercive methodologies persist primarily through the transmission of historical habit rather than proven necessity (Most, 1910/2001), peer-reviewed data underscores that force-free, relationship-based frameworks yield superior task-willingness, lower stress metrics, and secure attachment bonds (AVSAB, 2021; Hiby et al., 2004; Rooney & Cowan, 2011; Vieira de Castro et al., 2019, 2020). As advances in genetics, cognition, and neurobiology continue to refine animal management, the responsibility rests entirely upon policymakers, legislators, and handlers to align public policy with modern science, permanently retiring archaic, high-welfare-cost aversive tools in pursuit of a safer, more humane human-canine bond (Cook et al., 2016; Topál et al., 1998).
Albuquerque, N., Guo, K., Wilkinson, A., Savalli, C., Otta, E., & Mills, D. (2016). Dogs recognize dog and human emotions. Biology Letters, 12(1), 20150883. https://doi.org/10.1098/rsbl.2015.0883
American Society for the Prevention of Cruelty to Animals. (n.d.). Position statement on training aids and methods. https://www.aspca.org/about-us/aspca-policy-and-position-statements/position-statement-training-aids-and-methods
American Veterinary Society of Animal Behavior. (2021). Position statement on humane dog training. https://avsab.org/wp-content/uploads/2021/08/AVSAB-Humane-Dog-Training-Position-Statement-2021.pdf
American Veterinary Society of Animal Behavior. (2025, May 28). AVSAB board statement: Clarification on trainer endorsement and humane training methods. https://avsab.org/resources/position-statements/
Andics, A., Gácsi, M., Faragó, T., Kis, A., & Miklósi, Á. (2014). Voice-sensitive regions in the dog and human brain are revealed by comparative fMRI. Current Biology, 24(5), 574–578. https://doi.org/10.1016/j.cub.2014.01.058
Animal Welfare (Sentience) Act 2022, (c. 22) (UK). https://www.legislation.gov.uk/ukpga/2022/22/contents
Animal Welfare Act [Laki eläinten hyvinvoinnista], 693/2023 (Fin.). https://www.finlex.fi/en/legislation/2023/693
Animal Welfare Act [Dyrevelferdsloven], 2009-06-19-97 (Nor.). https://www.animallaw.info/article/norway-government-notes-2010-animal-welfare-act
Animal Welfare Act [Djurskyddslag], SFS 2018:1192 (Swed.). https://lagen.nu/2018:1192
Assemblée Nationale. (2023, January 10). Interdiction des colliers électriques et étrangleurs:
Adoption d’une proposition de loi [Prohibiting electric and choke collars: Adoption of a
bill]. République Française. https://assemblee-nationale.fr
Australian Veterinary Association. (2015, October 8). Warning against electronic collars. Vet Practice Magazine. https://www.vetpracticemag.com.au/warning-against-electronic-collars/
Australian Veterinary Association. (2021, July 22). Use of behaviour-modifying collars on dogs. https://www.ava.com.au/policy/use-of-behaviour-modifying-collars-on-dogs
Bekoff, M. (2007). The emotional lives of animals: A leading scientist explores animal joy, sorrow, and empathy, and why they matter. New World Library.
Bergström, A., Frantz, L., Schmidt, R., Ersmark, E., Lebrasseur, O., Girdland-Flink, L., Lin, A. T., Storå, J., Sjögren, K.-G., Anthony, D., Antipina, E., Amiri, S., Bar-Oz, G., Bazaliiskii, V. I., Bulatović, J., Brown, D., Carmagnini, A., Davy, T., Fedorov, S., . . . Skoglund, P. (2020). Origins and genetic legacy of prehistoric dogs. Science, 370(6516), 557–564. https://doi.org/10.1126/science.aba9572
Berns, G. S., Brooks, A. M., & Spivak, M. (2012). Functional MRI in awake unrestrained dogs. PLOS ONE, 7(5), e38027. https://doi.org/10.1371/journal.pone.0038027
Berns, G. S., Brooks, A. M., & Spivak, M. (2015). Scent of the familiar: An fMRI study of canine brain responses to familiar and unfamiliar human and dog odors. Behavioural Processes, 110, 37–46. https://doi.org/10.1016/j.beproc.2014.02.011
Berns, G. S., Brooks, A. M., Spivak, M., & Levy, K. (2017). Functional MRI in awake dogs predicts suitability for assistance work. Scientific Reports, 7, Article 43704. https://doi.org/10.1038/srep43704
Besluit van 14 juli 2021, houdende wijziging van het Besluit houders van dieren vanwege een aanpassing van het verbod op het gebruik van stroomstootapparatuur, Stb. 2021, 361. https://zoek.officielebekendmakingen.nl/stb-2021-361.html
Blackwell, E. J., Twells, C., Seawright, A., & Casey, R. A. (2008). The relationship between training methods and the occurrence of behavior problems, as reported by owners, in a population of domestic dogs. Journal of Veterinary Behavior, 3(5), 207–217. https://doi.org/10.1016/j.jveb.2007.10.008
Boletín Oficial del Estado. (2023). Ley 7/2023, de 28 de marzo, de protección de los derechos y el bienestar de los animales [Law 7/2023, of March 28, on the protection of the rights and welfare of animals]. Jefatura del Estado, (75). https://boe.es
British Columbia Society for the Prevention of Cruelty to Animals. (2016, July). Position statement on animal training. https://spca.bc.ca/programs-services/leaders-in-our-field/position-statements/position-statement-on-animal-training/
British Small Animal Veterinary Association. (2022). Position statement: Aversive training methods. https://www.bsava.com/position-statement/aversive-training-methods/
Bundesgesetz über den Schutz der Tiere (Tierschutzgesetz – TSchG) [Federal Act on the Protection of Animals], BGBl. I Nr. 118/2004, § 5 (Austria). https://www.ris.bka.gv.at/Dokumente/Erv/ERV_2004_1_118/ERV_2004_1_118.html
Bundesverwaltungsgericht. (2006). BVerwG 3 C 14.05. https://www.bverwg.de/230206U3C14.05.0
Buttner, A. P., Awalt, S. L., & Strasser, R. (2023). Early life adversity in dogs produces altered physiological and behavioral responses during a social stress-buffering paradigm. Journal of the Experimental Analysis of Behavior, 120(1), 6–20. https://doi.org/10.1002/jeab.856
Call, J., Bräuer, J., Kaminski, J., & Tomasello, M. (2003). Domestic dogs (Canis familiaris) are sensitive to the attentional state of humans. Journal of Comparative Psychology, 117(3), 257–263. https://doi.org/10.1037/0735-7036.117.3.257
Canadian Veterinary Medical Association. (2021, September 17). Humane training of dogs. https://www.canadianveterinarians.net/policy-and-outreach/position-statements/statements/humane-training-of-dogs/
Canadian Veterinary Medical Association. (2024, December 10). Canadian Association of Professional Dog Trainers code of ethics update. https://www.canadianveterinarians.net/about-cvma/latest-news/canadian-association-of-professional-dog-trainers-code-of-ethics-update/
Casey, R. A., Naj-Oleari, M., Campbell, S., Mendl, M., & Blackwell, E. J. (2021). Dogs are more pessimistic if their owners use two or more aversive training methods. Scientific Reports, 11, Article 19023. https://doi.org/10.1038/s41598-021-97743-0
China, L., Mills, D. S., & Cooper, J. J. (2020). Efficacy of dog training with and without remote electronic collars vs. a focus on positive reinforcement. Frontiers in Veterinary Science, 7, Article 508. https://doi.org/10.3389/fvets.2020.00508
Cook, P. F., Prichard, A., Spivak, M., & Berns, G. S. (2016). Awake canine fMRI predicts dogs' preference for praise vs food. Social Cognitive and Affective Neuroscience, 11(12), 1853–1862. https://doi.org/10.1093/scan/nsw102
Cooper, J. J., Cracknell, N., Hardiman, J., Wright, H., & Mills, D. (2014). The welfare consequences and efficacy of training pet dogs with remote electronic training collars in comparison to reward based training. PLOS ONE, 9(9), Article e102722. https://doi.org/10.1371/journal.pone.0102722
Consolidated Version of the Treaty on the Functioning of the European Union art. 13, 26 October 2012, OJ C 326, p. 47–390. europa.eu
Danish Animal Welfare Act [Dyreværnsloven], LBK nr 1 nr 20 af 11/01/2010 (Den.). https://www.retsinformation.dk/eli/lta/2010/20
Darwin, C. (1872). The expression of the emotions in man and animals. John Murray.
Deldalle, S., & Gaunet, F. (2014). Effects of 2 training methods on stress-related behaviors of the dog (Canis familiaris) and on the dog–owner relationship. Journal of Veterinary Behavior, 9(2), 58–65. https://doi.org/10.1016/j.jveb.2013.11.004
Eurogroup for Animals. (2024, November 12). EU animal welfare legislation review: Prioritising force-free standards across member states. https://www.eurogroupforanimals.org/news/eu-animal-welfare-legislation-review-prioritising-force-free-standards
Eurogroup for Animals. (2025, June 18). Advancing companion animal protection: A unified stance on banning electronic training devices. https://www.eurogroupforanimals.org/library/advancing-companion-animal-protection-unified-stance
Federation of Veterinarians of Europe, Federation of European Equine Veterinary Associations, Federation of European Companion Animal Veterinary Associations, & World Small Animal Veterinary Association. (2024, October 29). Joint position paper on the behaviour and training of animals. https://fve.org/publications/fve-feeva-fecava-and-wsava-joint-position-paper-on-the-behaviour-and-training-of-animals/
Finnish Kennel Club. (2024). The Animal Welfare Act protects the health and well-being of dogs. https://www.kennelliitto.fi/en/dog-ownership/animal-welfare-act-protects-health-and-well-being-dogs
Fugazza, C., Dror, S., Sommese, A., Temesi, A., & Miklósi, Á. (2021). Word learning dogs (Canis familiaris) provide an animal model for studying exceptional performance. Scientific Reports, 11, Article 14070. https://doi.org/10.1038/s41598-021-93581-2
Fugazza, C., Pogány, Á., & Miklósi, Á. (2016). Recall of others' actions after incidental encoding reveals episodic-like memory in dogs. Current Biology, 26(23), 3209–3213. https://doi.org/10.1016/j.cub.2016.09.057
Grundgesetz für die Bundesrepublik Deutschland [GG] [Basic Law], art. 20a (amended 2002). https://www.gesetze-im-internet.de/gg/art_20a.html
Hansen Wheat, C., van der Bijl, W., & Wynne, C. D. L. (2023). Rearing condition and willingness to approach a stranger explain differences in point following performance in wolves and dogs. Learning & Behavior, 51(2), 127–130. https://doi.org/10.3758/s13420-022-00544-2
Hare, B., Brown, M., Williamson, C., & Tomasello, M. (2002). The domestication of social cognition in dogs. Science, 298(5598), 1634–1636. https://doi.org/10.1126/science.1072702
Hare, B., & Tomasello, M. (2005). Human-like social skills in dogs? Trends in Cognitive Sciences, 9(9), 439–444. https://doi.org/10.1016/j.tics.2005.07.003
Haverbeke, A., Laporte, B., Depiereux, E., Giffroy, J.-M., & Diederich, C. (2008). Training methods of military dog handlers and their effects on the team's performances. Applied Animal Behaviour Science, 113(1–3), 110–122. https://doi.org/10.1016/j.applanim.2007.11.010
Herron, M. E., Shofer, F. S., & Reisner, I. R. (2009). Survey of the use and outcome of confrontational and non-confrontational training methods in client-owned dogs showing undesired behaviors. Applied Animal Behaviour Science, 117(1–2), 47–54. https://doi.org/10.1016/j.applanim.2008.12.011
Hiby, E. F., Rooney, N. J., & Bradshaw, J. W. S. (2004). Dog training methods: Their use, effectiveness and interaction with behaviour and welfare. Animal Welfare, 13(1), 63–69. https://doi.org/10.1017/S0962728600026683
Irish Society for the Prevention of Cruelty to Animals. (n.d.). Responsible pet ownership and positive dog training principles. https://www.ispca.ie/advice/responsible_pet_ownership/positive_dog_training
Iceland Ministry of Industries. (2013). Act No. 55/2013 on animal welfare. Government of Iceland. https://government.is/publications/legislation/act/2018-06-05-Act-No.-55-2013-on-animal-welfare
International Association of Canine Professionals. (2025, August 18). Tracking legislation worldwide. https://iacpdogs.org/2025/08/18/%F0%9F%8C%8D-tracking-legislation-worldwide/
Janssens, L., Giemsch, L., Schmitz, R., Street, M., Van Dongen, S., & Crombé, P. (2018). A new look at an old dog: Bonn-Oberkassel reconsidered. Journal of Archaeological Science, 92, 126–138. https://doi.org/10.1016/j.jas.2018.01.004
Kaminski, J., Call, J., & Fischer, J. (2004). Word learning in a domestic dog: Evidence for “fast mapping.” Science, 304(5677), 1682–1683. https://doi.org/10.1126/science.1097859
Kennel Union of South Africa. (2024, January 15). Notice to affiliated clubs, provincial councils and national sub-committees re: Use of training dog collars. https://www.kusa.co.za/index.php/news/kusa-news/3180-notice-to-affiliated-clubs-provincial-councils-and-national-sub-committees-re-use-of-training-dog-collars
Miklósi, Á., Polgárdi, R., Topál, J., & Csányi, V. (1998). Use of experimenter-given cues in dogs. Animal Cognition, 1(2), 113–121. https://doi.org/10.1007/s100710050016
Ministère de l'Agriculture et de la Souveraineté alimentaire. (2025, June 19). Arrêté du 19 juin
2025 fixant les règles sanitaires et de protection animale auxquelles doivent satisfaire les
activités liées aux animaux de compagnie d'espèces domestiques [Decree of June 19, 2025
establishing health and animal protection rules for activities related to domestic
companion animals]. Journal Officiel de la République Française.
Morrill, K., Hekman, J., Li, X., McClure, J., Logan, B., Goodman, L., Gao, M., Dong, Y., Alonso, M., Carmichael, E., Snyder-Mackler, N., Alonso, J., Noh, H. J., Johnson, J., Koltookian, M., Lieu, C., Megquier, K., Swofford, R., Turner-Maier, J., . . . Karlsson, E. K. (2022). Ancestry-inclusive dog genomics challenges popular breed stereotypes. Science, 376(6592), Article eabk0639. https://doi.org/10.1126/science.abk0639
Most, K. (2001). Training dogs: A manual. Dogwise Publishing. (Original work published 1910)
Oesterreich.gv.at. (2025, March 27). General information on keeping dogs in a species-appropriate manner. Federal Chancellery of Austria. Retrieved August 30, 2026, from https://www.oesterreich.gv.at/en/themen/reisen_und_freizeit/haustiere/1/1/Seite.742110
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
Ovodov, N. D., Crockford, S. J., Kuzmin, Y. V., Higham, T. F. G., Hodgins, G. W. L., & van der Plicht, J. (2011). A 33,000-year-old incipient dog from the Altai Mountains of Siberia: Evidence of the earliest domestication disrupted by the Last Glacial Maximum. PLOS ONE, 6(7), Article e22821. https://doi.org/10.1371/journal.pone.0022821
Pavlov, I. P. (1927). Conditioned reflexes: An investigation of the physiological activity of the cerebral cortex (G. V. Anrep, Ed. & Trans.). Oxford University Press.
Perri, A. R., Feuerborn, T. R., Frantz, L. A. F., Larson, G., Malhi, R. S., Meltzer, D. J., & Witt, K. E. (2021). Dog domestication and the dual dispersal of people and dogs into the Americas. Proceedings of the National Academy of Sciences, 118(6), Article e2010083118. https://doi.org/10.1073/pnas.2010083118
Pilley, J. W., & Reid, A. K. (2011). Border collie comprehends object names as verbal referents. Behavioural Processes, 86(2), 184–195. https://doi.org/10.1016/j.beproc.2010.11.007
Protocol on protection and welfare of animals, annexed to the Treaty of Amsterdam amending the Treaty on European Union, the Treaties establishing the European Communities and certain related acts, 2 October 1997, OJ C 340, 10.11.1997, p. 110. https://eur-lex.europa.eu/eli/treaty/ams/pro_10/sign/eng
Rooney, N. J., & Cowan, S. (2011). Training methods and owner–dog interactions: Links with dog behaviour and learning ability. Applied Animal Behaviour Science, 132(3–4), 169–177. https://doi.org/10.1016/j.applanim.2011.03.007
Royal New Zealand Society for the Prevention of Cruelty to Animals. (2024, April). Position statement on training methods and devices. https://www.spca.nz/advocacy/position-statements/article/training-methods-and-devices
Royal Society for the Prevention of Cruelty to Animals Australia. (2023, February 8). Is the use of electronic dog collars legal? https://kb.rspca.org.au/categories/legislation/companion-animal-legislation/is-the-use-of-electronic-dog-collars-legal
Royal Society for the Prevention of Cruelty to Animals Australia. (2025, December 14). Is it important to train my dog? What sort of training would you recommend? https://kb.rspca.org.au/categories/companion-animals/dogs/training/is-it-important-to-train-my-dog-what-sort-of-training-would-you-recommend
Royal Society for the Prevention of Cruelty to Animals Australia. (2026, January 21). RSPCA Policy C7 Training, behaviour modification and invasive procedures. https://kb.rspca.org.au/categories/rspca-policy/c-animals-in-sport-entertainment-performance-recreation-and-work/rspca-policy-c7-training-behaviour-modification-and-invasive-procedures
Royal Society for the Prevention of Cruelty to Animals UK. (n.d.-a). Help ban the use of shock collars. https://www.rspca.org.uk/getinvolved/campaign/shockcollarsRoyal Society for the Prevention of Cruelty to Animals UK. (n.d.-b). Why prong collars are harmful to dogs. https://www.rspca.org.uk/adviceandwelfare/pets/dogs/training/prongcollars
RSPCA Australia. (2023a, March 3). What are the animal welfare issues with electric shock collars on dogs? RSPCA Knowledgebase. https://kb.rspca.org.au/categories/companion-animals/dogs/training/what-are-the-animal-welfare-issues-with-electric-shock-collars-on-dogs
RSPCA Australia. (2023b, February 8). Is the use of electronic dog collars legal? RSPCA Knowledgebase. https://kb.rspca.org.au/categories/legislation/companion-animal-legislation/is-the-use-of-electronic-dog-collars-legal
Salomons, H., Smith, K. C. M., Callahan-Beckel, M., Callahan, M., Levy, K., Kennedy, B. S., Bray, E. E., Gnanadesikan, G. E., Horschler, D. J., Gruen, M., Tan, J., White, P., vonHoldt, B. M., MacLean, E. L., & Hare, B. (2021). Cooperative communication with humans evolved to emerge early in domestic dogs. Current Biology, 31(14), 3137–3144. https://doi.org/10.1016/j.cub.2021.06.051
Satellai. (2026, July 22). E-collar laws by country and state: Avoid big fines. Satellai Pet Tech. https://satellai.com/blogs/tech-enabled-pet-life/shock-collar-bans-map
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
Scott, J. P., & Fuller, J. L. (1965). Genetics and the social behavior of the dog. University of Chicago Press.
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
Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. Appleton-Century.
Starvey Global. (2026, August 23). EU & US shock collar regulations 2026 | Wholesale distributors. https://starveyglobal.com
Stiftung für das Tier im Recht. (2021, May 10). Illegal dog training aids. TIR Legal Center. https://www.tierimrecht.org/en/news/news-2021/2021-05-10-illegal-dog-training-aids/
The Animal Welfare (Electronic Collars) (Wales) Regulations 2010, S.I. 2010/943 (W. 97). https://www.legislation.gov.uk/wsi/2010/943/contents/made
Tierschutzgesetz [TierSchG] [Animal Welfare Act], v. 18.5.2006, BGBl. I S. 1206, 1313, zuletzt geändert durch Art. 2 Abs. 20 G v. 20.12.2022, BGBl. I S. 2752, § 3 Nr. 11. https://www.gesetze-im-internet.de/tierschg/
Topál, J., Miklósi, Á., Csányi, V., & Dóka, A. (1998). Attachment behavior in dogs (Canis familiaris): A new application of Ainsworth's (1969) Strange Situation Test. Journal of Comparative Psychology, 112(3), 219–229. https://doi.org/10.1037/0735-7036.112.3.219
Vie-Publique. (2023, January 17). Proposition de loi visant à interdire l'utilisation des colliers
étrangleurs et électriques pour les chiens et les chats [Bill aiming to ban the use of choke
and electric collars for dogs and cats]. Direction de l'information légale et
administrative. https://www.vie-publique.fr/loi/287859-loi-interdiction-colliers-etrangleurs-et-electriques-chiens-chats
Vier Pfoten Österreich. (2024, March 20). Quälende halsbänder [Tormenting collars]. FOUR PAWS International. https://www.vier-pfoten.at/unsere-geschichten/publikationen/quaelende-halsbaender
Vieira de Castro, A. C., Araújo, Â., Fonseca, A., & Olsson, I. A. S. (2021). Improving dog training methods: Efficacy and efficiency of reward and mixed training methods. PLOS ONE, 16(2), Article e0247321. https://doi.org/10.1371/journal.pone.0247321
Vieira de Castro, A. C., Barrett, J., de Sousa, L., & Olsson, I. A. S. (2019). Carrots versus sticks: The relationship between training methods and dog-owner attachment. Applied Animal Behaviour Science, 219, Article 104831. https://doi.org/10.1016/j.applanim.2019.104831
Vieira de Castro, A. C., Fuchs, D., Morello, G. M., Pastur, S., de Sousa, L., & Olsson, I. A. S. (2020). Does training method matter? Evidence for the negative impact of aversive-based methods on companion dog welfare. PLOS ONE, 15(12), Article e0225023. https://doi.org/10.1371/journal.pone.0225023
Welsh Government. (2017). Electronic collars in dogs and cats: A review of the welfare implications. https://www.gov.wales/sites/default/files/publications/2017-12/electronic-collars-in-dogs-and-cats-review-of-welfare-implications.pdf
Zakon o zaščiti živali [Animal Protection Act], Uradni list RS, št. 98/1999–109/2023 (Slov.).
https://leap.unep.org/en/countries/si/national-legislation/animal-protection-act-0
Ziv, G. (2017). The effects of using aversive training methods in dogs—A review. Journal of Veterinary Behavior, 19, 50–60. https://doi.org/10.1016/j.jveb.2017.02.004
© International Canine Legislation Action Committee. This document may be shared for advocacy and educational purposes with attribution.