Fagen et al. (elephant learning)

Fagen et al. (2014) is a field study showing that operant conditioning can train working elephants in Nepal to cooperate voluntarily with a stressful medical procedure. Using reward alone, researchers taught five elephants to offer their trunks for a tuberculosis test, replacing the punishment-based methods traditionally used to control them.

Psychology Being Investigated

1. Operant Conditioning

Operant conditioning is a theory of learning where behavior is influenced by its consequences.

Behavior that is reinforced (rewarded) will likely be repeated, and behavior that is punished will occur less frequently.

The study demonstrates that elephants can learn complex sequences of behaviors through positive reinforcement and that this learning can occur in natural settings.

This idea goes back further. Thorndike’s (1911) law of effect first proposed that a rewarded behaviour is “stamped in” and becomes more likely, while a punished one is “stamped out.” Skinner (1938) took this further.

He defined operant conditioning as how consequences change how often a voluntary behaviour happens again. Consequences that strengthen a behaviour are called reinforcement; consequences that weaken it are called punishment. Fagen et al. (2014) relies on just one type throughout: positive reinforcement.

2. Secondary Positive Reinforcement

Positive reinforcement means delivering something rewarding immediately after a behaviour, so the behaviour becomes more likely to happen again. It is the mechanism this whole study rests on.

A primary reinforcer satisfies a basic need on its own. Chopped bananas served as the primary reinforcer here, since they satisfy an elephant’s hunger directly.

A secondary reinforcer has no built-in value. It gains power only through classical conditioning.

A neutral stimulus, like a whistle, is repeatedly paired with a primary reinforcer, like bananas, until it starts to signal the same reward on its own. Secondary positive reinforcement, SPR, is training built on exactly this kind of paired-up reinforcer.

The whistle in Fagen et al.’s research became a secondary reinforcer this way. It was paired repeatedly with chopped bananas.

Skinner (1953) argued that reinforcement like this is a more precise tool for building new behaviour than punishment. Punishment only suppresses a response; it does not teach an alternative.

This is why the study contrasts so sharply with the punishment- and aversion-based methods traditionally used to train elephants in Nepal.

3. Training Methods & Animal Welfare

Humane training methods like SPR can positively affect animal mental health and welfare, reducing stress and improving cooperation.

This is particularly relevant here. The study trains elephants to voluntarily participate in a trunk wash for tuberculosis testing, a procedure that could be stressful under traditional training methods.

SPR succeeded at training elephants to voluntarily take part in a potentially stressful veterinary procedure. By demonstrating this, the study advocates for a more humane and ethical approach to animal training and care.

Background

Traditional elephant training in Nepal heavily relies on punishment and aversion-based techniques, often through mahouts using physical control and restraint.

That carries real risk. Restraining a multi-tonne animal by force is hazardous to handlers and stressful for the elephant.

However, there has been a global shift towards more humane training approaches, specifically positive reinforcement.

The stakes are real.

Fagen et al. wanted to investigate whether secondary positive reinforcement (SPR) could train captive elephants to voluntarily participate in a trunk wash.

A trunk wash is the standard procedure for testing elephants for tuberculosis: sterile saline is flushed into the trunk, then blown out into a container for laboratory testing.

This matters for two species at once. TB is a significant, zoonotic health concern, transmissible between elephants and the people who work closely with them in camps and tourism.

Aim

The primary aim was to investigate whether secondary positive reinforcement (SPR) techniques could effectively train traditionally trained elephants in Nepal to voluntarily participate in trunk wash procedures.

More broadly, the study aimed to show that reward-based, cooperative husbandry is a viable, humane alternative to traditional punishment-based elephant training. It also aimed to document a method other handlers could reproduce, improving both their own safety and the elephants’ welfare.

Research Method

The study employed an observational research method.

This means researchers systematically observed and recorded the elephants’ behavior during the training process without directly manipulating other variables in the environment.

This approach allowed for an in-depth understanding of the elephants’ responses to the training methods and how their behavior changed over time.

Sample

The study involved a small sample of five female Asian elephants: four juveniles, aged 5 to 7, and one adult, estimated to be in her 50s.

All five were housed at the same stable in Nepal.

The researchers used opportunity sampling to select them.

This means they picked elephants that were readily available and met set criteria. Each had to be docile, not pregnant or nursing a calf, and have a mahout willing to take part.

Experimental Design

The researchers used a pre-post design with repeated measures.

This design collects data on the elephants’ behaviour before the training intervention (pre-training). It then collects data again after the training programme is complete (post-training) to assess what changed.

The repeated measures aspect means that the same elephants were assessed at both time points, allowing for direct comparisons of individual elephant’s behavior before and after the SPR training.

Independent Variable (IV)

The independent variable, which is the factor being manipulated or changed by the researchers, was the implementation of the SPR training technique.

The researchers controlled the introduction and progression of the SPR training steps, making it the variable they systematically altered to observe its effects on the elephants’ behavior.

Dependent Variables (DV)

The study had multiple dependent variables, which are the factors being measured or observed to assess the effects of the independent variable. These included:

• Training session duration: The assistant timed the session to the nearest minute. It started from when the first cue was offered. It ended after the elephant’s response to the last cue offered.
• Number of offers: Researchers recorded how many times the elephants successfully offered their trunks through the hole in the wall.
• Percentage of correct responses: Performance tests were conducted every five training sessions to assess the elephants’ mastery of the tasks. Researchers calculated the percentage of correct responses or successful completion of each behavioral task during these tests. This provided a quantitative measure of the elephants’ learning progress and the effectiveness of the training in shaping their behavior.
• Cumulative time spent on training: The total time in minutes spent training each elephant throughout the study. The measurement involved summing the duration of all training sessions for each individual elephant.
• Overall improvement in performance: Researchers analyzed the progression of learning throughout the study by looking at the mean percent correct in all tasks for all elephants for each test session. This measure reflects the overall trend of skill acquisition and improvement in the elephants’ performance as a result of the SPR training.

Procedure

The study followed a systematic procedure to implement and evaluate the SPR training:

1. Selection Criteria:

Elephants were selected based on their docility, availability (not pregnant or nursing), and the willingness of their mahouts to participate in the study.

2. Training Methodology:

The training exclusively used the SPR technique, with chopped bananas serving as the primary reinforcer and a short whistle blow as the secondary reinforcer.

The whistle was initially a neutral stimulus, but it became associated with the primary reinforcer (chopped bananas) through classical conditioning.

Over time, the elephants learned to associate the whistle with the reward, making the whistle itself a powerful motivator for performing the desired trunk movements

• Primary reinforcer: Chopped bananas, which directly satisfy the elephant’s hunger.
• Secondary reinforcer: A short whistle blast, paired with the bananas to create a conditioned association. Elephants learn to associate the whistle with the reward (bananas), so it becomes a powerful reinforcer on its own.

3. Training Setting:

Training sessions were conducted in the mornings (7:30-10:00 am) and afternoons (4:00-7:00 pm) while the elephants were chained in their stalls.

The mahouts, the elephants’ handlers, were present for safety but were instructed not to interact with the elephants during training to maintain the integrity of the training.

This ensured that any observed changes were solely attributable to the SPR training.

To ensure the elephants would not drink the saline solution or water used for the trunk wash, they were offered water before each session.

4. Training Techniques:

The elephants next went through a progressive series of training stages using secondary positive reinforcement techniques.

The elephants were then trained to perform these tasks in sequence: behavioural chaining. Each step became the cue for the next. This let the whole routine run end to end, culminating in the full trunk wash sequence.

The researchers adopted a flexible approach to training, tailoring the plan to each elephant’s individual needs and learning pace.

There was no fixed time limit for each stage; progression was determined by the elephant’s success and comfort level. This individualized approach likely contributed to the positive training outcomes.

1. Bridge Training: Establishing the link between the primary and secondary reinforcer by repeatedly pairing the whistle with the banana reward.
2. Target Behaviors: The elephants were then trained individually on five basic behavioral tasks essential for the trunk wash:
   • Trunk Here: The elephant places the tip of her trunk into the trainer’s hand.
   • Trunk Up: The elephant lifts her trunk upwards.
   • Bucket: The elephant places her trunk tip into a bucket.
   • Blow: The elephant blows air forcefully out of her trunk.
   • Steady: Maintaining a specific trunk position for a certain duration.
3. Individual Behavior Training: The researchers trained each of the five target behaviors separately using:
   • Capture: Involves reinforcing a behavior that the animal already spontaneously performs, shaping it towards the desired action. The researchers used capturing to train the elephants to “blow,” or exhale forcefully through their trunks. They waited for the elephants to exhale naturally and then marked and rewarded the behavior with a banana.
   • Lure: Uses a reward to guide the animal into the desired position or action. Luring was employed in several tasks, including “trunk here” and “trunk up.” For “trunk here,” the trainer initially placed chopped bananas in their hand, encouraging the elephant to place its trunk on the trainer’s palm to retrieve the reward. For “trunk up,” the trainer lifted the banana above the elephant’s head, prompting it to lift its trunk to reach the treat. In both cases, the lure was gradually faded as the elephant learned the behavior.
   • Shaping: Gradually reinforces successive approximations of the desired behavior, rewarding increasingly closer attempts until the full behavior is achieved. For instance, in the “blow” task, the researchers initially rewarded any exhale directed towards the bucket. As the elephants became more proficient, they only rewarded exhales aimed at the center of the bucket, gradually shaping the behavior towards the desired accuracy.
4. Verbal Cues: Once an elephant mastered a behavior, a unique, monosyllabic verbal cue was paired with the action. These cues were non-words to avoid any potential understanding of the cues by the mahouts. The verbal cues played a crucial role in the behavioral chaining process. As the elephants learned to associate each cue with a specific behavior, the trainers could then use the cues to prompt the elephants through the complete trunk wash sequence. The completion of one behavior, triggered by its corresponding verbal cue, would then serve as a cue for the next behavior in the chain.
5. Behavioral Chaining: After mastering individual behaviors, the elephants were trained to link them together in sequence. For the trunk wash, this involved performing “Trunk Here,” “Trunk Up,” “Bucket,” and “Blow” in a smooth, continuous sequence.
   • Short Sequences: Initially, trainers started with short sequences, such as combining “bucket” and “blow.”
   • Longer Sequences: Gradually, these sequences were expanded until the elephants could perform the entire trunk wash procedure: “trunk here,” “steady,” “trunk up,” “steady,” “bucket,” and “blow.”
   • Rewarding the Complete Sequence: The elephants received the primary reinforcer (banana) only upon successfully completing the entire sequence in the correct order. This motivated them to link the behaviors together and perform the trunk wash as a unified action.
6. Syringe and Fluid Desensitization: This crucial stage involved acclimating the elephants to the syringe and the sensation of fluid being inserted into their trunks. The researchers achieved this using:
   • Desensitization: Gradually exposing the elephants to the syringe, starting by touching it to the outside of the trunk and progressing to inserting it into the nostril.
   • Counterconditioning: Pairing the presentation of the syringe with the banana reward to create a positive association and reduce aversion.
   • Gradual Fluid Introduction: Initially injecting only a drop of fluid and slowly increasing the volume in small increments until the full 60ml needed for sample collection was tolerated.

5. Data Collection:

Detailed data was collected throughout the training process:

• Session Times: Timed the same way described under Dependent Variables above.
• Number of Offers: Recorded to track the elephants’ responsiveness to cues for each behavior.
• Performance Tests: Administered every 5 sessions, starting at Session 10, to assess mastery of individual behaviors and sequences. A passing score was 80% or higher. If an elephant failed, they were retested to pinpoint areas for improvement.

Results

• Success Rate: Out of the five elephants, four, all juveniles, successfully learned the complete trunk wash sequence. The adult elephant, despite making progress, did not learn the entire sequence.
• Training Duration: The total training time varied between the elephants. The shortest training time was 257 minutes, and the longest was 451 minutes. This difference could be due to individual learning paces and variations in mastering specific tasks.
• Task Difficulty: The researchers found that the “Trunk Here” task, where the elephant had to place the tip of her trunk into the trainer’s hand, proved to be the most challenging and required the most practice.

Conclusions

• SPR as an Effective Training Method: The study demonstrated that SPR can be successfully implemented to train elephants for veterinary procedures like the trunk wash. This method, which relies on positive reinforcement and gradual desensitization, proved effective in teaching elephants a sequence of behaviors necessary for sample collection.
• Implications for Elephant Welfare: The study highlighted the potential benefits of SPR for elephant welfare. By using positive reinforcement, the training process can be less stressful and more humane compared to traditional elephant training techniques in Nepal, which often involve punishment and aversion. Voluntary participation in veterinary procedures can also reduce stress and improve the safety of both the elephants and their handlers.
• Age as a Factor in Training Success: The study observed that all four juvenile elephants learned the trunk wash, while the adult elephant did not. This finding suggests that age may play a role in an elephant’s learning ability and that younger elephants might be more receptive to this type of training.
• Future Applications: The success of this study opens up possibilities for using SPR to train elephants for other essential healthcare procedures. This could lead to improvements in elephant management and veterinary care, ultimately contributing to better overall health and well-being for captive elephants.

Strengths

Several methodological choices strengthen this study’s conclusions:

1. Standardisation and Internal Validity: Every elephant learned the same five behaviours in the same order, and the mahouts stayed silent during sessions so only the researchers’ training shaped the results.
2. Objective Measurement: Countable measures, training minutes, number of successful responses, percentage correct, made the findings replicable and open to scrutiny.
3. Ecological Validity: Training took place in the elephants’ own working environment, so the behaviour observed is more likely to reflect real husbandry settings.
4. Replicability: The detailed, step-by-step methodology can be reproduced by other trainers and camps working with elephants elsewhere.
5. Ethical Safeguards: The whole design was built around voluntary participation, with elephants free to withdraw at any point.

Standardisation and Internal Validity

The researchers used standardised training procedures across all five elephants. Each one learned the same five behaviours, Trunk Here, Trunk Up, Bucket, Blow and Steady, in the same order.

The mahouts normally interact with the elephants all day. During sessions, though, they were told to stay silent.

That single rule closes off an obvious alternative explanation. Without it, a sceptic could argue the elephants were simply obeying a familiar voice, not learning something new from the whistle.

Together, these controls strengthen the study’s internal validity. Any observed differences in learning are best explained by the elephants themselves, not by inconsistent training or stray cues.

This is standard practice in behavioural research. Hold everything constant except the one thing you want to test.

Objective Measurement

Learning was scored with countable, comparable measures. These included training minutes, the number of successful responses, and the percentage of correct responses at each performance test.

A fixed 80% mastery criterion decided when a behaviour counted as learned. That is a clear line, not a matter of opinion.

This kind of quantitative data reduces reliance on a trainer’s subjective impression. Session times and number of offers let the researchers compare elephants objectively, rather than by feel.

In the Fagen study, this meant every elephant’s progress could be tracked on the same numeric scale. It was not just described in general terms.

This is not a small thing. Subjective judgement is exactly where evaluations of animal training usually go wrong, and this study largely avoids it.

Ecological Validity

Training happened in the elephants’ own working environment. Their usual handlers were present throughout.

The researchers even noted the natural sounds of the jungle as part of that setting. Nothing about it was artificial.

Because the study was run this way, the behaviour observed is more likely to reflect how the technique performs in real husbandry settings. This includes other camps, safari parks and zoos, not just a laboratory.

That matters for anyone hoping to apply the technique elsewhere. It was never tested only under artificial, controlled conditions.

This matters practically too. A method proven in a real camp, with real distractions and a real handler relationship, is more trustworthy than one proven in a laboratory alone. Laboratory conditions rarely resemble how elephants actually live and work.

Replicability

The researchers documented their methodology in detail. This covered the target behaviours, the reinforcers used, the training techniques, and the criteria for mastering each step.

Nothing about the procedure was left vague. Every stage is spelled out.

That level of detail means other trainers and camps could reproduce the programme with a different group of elephants, letting the findings be checked rather than simply taken on trust.

A method that only works for its original researchers is far less useful. One that any competent trainer could pick up and run is worth far more.

This is not just a nice bonus. A finding that only one team, in one setting, can produce is much weaker evidence than one that holds up wherever a competent trainer tries it.

Ethical Safeguards

The study demonstrated a strong commitment to ethical guidelines, built around voluntary participation throughout.

Withdrawal. Elephants could leave a training session at any time simply by walking away. Nothing forced them to stay. This respected their autonomy and reduced any distress the training might otherwise cause. It also kept trainers safer: a coerced elephant is a more dangerous one.

Minimizing potential distress. The trunk wash was paired with positive experiences throughout: chopped bananas and the whistle’s sound. This encouraged the elephants to cooperate willingly rather than through fear. Reward replaced coercion completely.

The approach fits humane-training principles more broadly, using reward as a more ethical alternative to punishment or negative reinforcement.

Elephant 1, the youngest, showed signs of impatience in longer sessions, so researchers shortened hers. Every elephant is different.

Handlers also offered water before each session, so the elephants would not drink the saline used for the trunk wash. Small adjustments like these show real attentiveness to each elephant’s needs, not just to the training schedule.

Weaknesses

The study also has real limitations worth weighing against its strengths:

1. Small, Unrepresentative Sample: Only five elephants, all female, from a single Nepalese stable, selected for docility and availability, limits how far the findings generalise.
2. No Control Condition: There was no comparison group trained by traditional methods, or left untrained, so the study cannot show how much better SPR is than the alternative.
3. Subjective Scoring: Whether a shaped behaviour counted as “correct” was ultimately a trainer’s judgement call, which could bias the reported success rate.
4. Inflated Training Times: Total training time included practice on behaviours later dropped from the final sequence, making the reported times look longer than necessary.
5. Uncontrolled Distractions: Tourists, other animals and camp activity were present during training and could not be controlled for.
6. No Follow-Up: The study never checked whether the elephants still performed the trunk wash reliably weeks or months later.

Small, Unrepresentative Sample

The study included just five female Asian elephants, all from a single stable in Nepal. They were selected because they were docile, available, and had willing handlers.

That is a small, narrow sample.

This raises real doubts about how far the results generalise. The findings might not apply to male elephants, older elephants, or elephants from other regions with different training histories.

A larger, more diverse sample would strengthen these conclusions and give more robust evidence for SPR across elephant populations. One case stands out. The sole adult elephant failed to learn the full sequence.

The researchers suggest adult elephants may need a different training approach, or simply more time to adjust. Confirming that is a question for future research.

No Control Condition

The study never included a comparison group. No elephants were trained by traditional, punishment-based methods for comparison, and none were left completely untrained.

Without that kind of control, the study cannot directly quantify how much better SPR is than the alternative.

It demonstrates that the method works. It does not demonstrate that the method outperforms traditional training under matched conditions.

A control group would have shown what happens without reward-based training under otherwise identical conditions. Without one, some of the improvement could in principle reflect simple familiarity with the routine over time, not the reinforcement itself.

This is a common trade-off in applied field research. Testing a new method on animals already destined for real training work leaves little room for an untrained control group. That gap still limits what the study can conclude.

Subjective Scoring

Judging whether a shaped behaviour counted as correct was ultimately a trainer’s call. That leaves room for bias.

The reported success rate could have been inflated by this subjectivity. Determining the exact angle of the trunk during the “steady” task, for instance, is a judgement call. Different observers might read it differently.

Video recordings reviewed by multiple observers would help. That single change could remove much of the guesswork.

As it stands, the study’s precise success figures deserve a little scepticism. They rest on one trainer’s eye, not an independently checked measure.

This is not unique to this study. Any research relying on human judgement calls faces the same risk, which is exactly why more objective measures are valuable wherever they are possible.

Inflated Training Times

The total training time included practice on behaviours later dropped from the final sequence. Four elephants, for example, were initially trained on a “trunk-out” behaviour that turned out to be unnecessary.

Counting that time inflates how long the essential trunk-wash training actually took. That is misleading.

Excluding it would give a more accurate measure of training efficiency. It also matters for anyone using the reported times to plan their own programme.

An inflated figure makes SPR look slower than it really is. This is a small methodological point.

But it matters for anyone trying to replicate or compare the results. A training log that mixes essential and non-essential practice makes the numbers harder to interpret than one that reports only the behaviours that made it into the final routine.

Uncontrolled Distractions

Training happened where tourists, other animals and camp activity were all present. These are real, uncontrolled variables.

The study notes that the adult elephant was particularly distracted by a calf in a neighbouring stall. A more controlled setting, or active steps to reduce distractions, could have removed this noise from the results.

This is the usual trade-off of field research. Realism is gained, but at the cost of the tight control a laboratory study would have offered.

This is not a minor caveat. A training method that only works in a hushed, distraction-free setting would be far less useful than one that holds up amid the noise of a real working camp.

Testing it under real conditions is, in that sense, a strength as much as a weakness.

No Follow-Up

The study never checked whether the elephants still performed a reliable trunk wash weeks or months after training ended. That is a real gap.

Durability matters here: a technique that fades once intensive training stops has far less practical value than one that lasts. Future work could test retention at set intervals after training finishes to settle the question.

This is a common blind spot in single-site field studies. Measuring learning is easier than measuring maintenance.

Until retention is tested, SPR’s case rests on how well elephants learn the behaviour. It does not yet show how long they keep it, which is the more practically important question for working elephants.

Trainers considering this method for their own elephants should keep that open question in mind.

Comparison With Free-Contact and Protected-Contact Management

Fagen et al.’s trunk-wash programme sits inside a wider debate about how captive elephants should be handled at all.

Setting it against the field’s two established management systems sharpens what the study does, and does not, show.

Free-Contact vs Protected-Contact Management

Free-contact (FC) management is the traditional model, keeping mahout or keeper and elephant in the same space with no barrier between them. Compliance has historically relied on physical restraint and negative reinforcement, often through an ankus, a pointed hook applied to sensitive pressure points.

Protected-contact (PC) management separates keeper and elephant behind a barrier. Coercion becomes physically impossible: the elephant can only be worked with through positions it chooses to offer, so compliance has to be built entirely through reward.

Most accredited North American zoos adopted PC from the 1990s.

Neither system is inherently right or wrong on its own. Each involves a different trade-off between control, safety for handlers, and the animal’s freedom to decline a request entirely, and the right choice can depend on local resources.

Wilson et al. (2015): Comparing Reinforcement Rates

One study puts numbers on that trade-off directly.

This is a small, single comparison, not a randomised trial, but it gives a rare, direct empirical picture of how much aversive control traditional free-contact handling actually involves.

A Two-Decade Follow-Up

A 2026 follow-up from the same research group returned to the original zoo nearly twenty years after its move from free contact to protected contact. Even after a full rebuild, compliance stayed high.

That durability suggests the welfare gains from protected contact are a lasting feature of the training relationship, not a short-lived novelty. The 2026 study is itself small and single-site, so it corroborates the welfare case rather than proving it outright.

None of this is a randomised trial. All of it points the same way.

This does not settle the debate between the two systems. It does show that consistency over time matters just as much as which system a facility originally chooses, and that a well-run system can hold up for decades.

Issues and Debates

Application to everyday life: Training Animals

The study’s success in training elephants for a specific procedure, the trunk wash, highlights the potential of SPR in teaching animals a wide range of tasks and behaviors.

This has practical implications for various fields, including:

• Veterinary Care: Training animals to cooperate with medical procedures, such as vaccinations or examinations, can significantly reduce stress for both the animal and the veterinarian.
• Service Animal Training: Training service dogs to perform specific tasks to assist individuals with disabilities requires consistent and effective reinforcement techniques.
• Animal Enrichment in Zoos and Sanctuaries: SPR can be utilized to encourage natural behaviors and mental stimulation in captive animals, enhancing their overall well-being.

Application to everyday life: Human Learning and Behavior Change

• Parenting and Child Development: Positive reinforcement techniques, similar to those used with the elephants, can be highly effective in shaping desired behaviors in children. By consistently rewarding positive actions and providing clear expectations, parents and caregivers can foster pro-social behaviors, encourage responsibility, and motivate learning.
• Education and Learning: The study’s findings support the application of reinforcement principles in educational settings. Educators can utilize positive reinforcement strategies to motivate students, encourage participation, and promote academic achievement. This could involve:
  • Providing verbal praise and encouragement for effort and progress.
  • Implementing reward systems, such as points or privileges, for completing assignments or achieving learning goals.
  • Creating a positive learning environment that fosters a sense of accomplishment and motivates students to strive for success.
• Personal Growth and Habit Formation: The principles of reinforcement can be applied to personal growth and habit formation. By setting clear goals, identifying desired behaviors, and implementing a consistent system of rewards, individuals can increase the likelihood of achieving personal objectives, such as:
  • Developing healthy habits (exercise, balanced diet).
  • Improving time management skills.
  • Breaking unwanted habits (procrastination, unhealthy snacking).

The same principle scales up in an even more direct way. This is a familiar idea.

Token economies used in classrooms, hospitals and rehabilitation programmes work by exactly this logic. A token, itself worthless, is earned for a target behaviour and later exchanged for a genuine reward (Ayllon & Azrin, 1968).

The parallel is exact.

A sticker chart for a child and a whistle for an elephant are the same psychological device. Both are secondary reinforcers that make reward practical when the real prize cannot be handed over the instant the behaviour happens.

Nature versus nurture

While all elephants received the same standardised training, they showed individual differences in learning speed and performance.

That points to nature. These variations could reflect inherent differences in personality, cognitive ability, or prior experience.

Elephant 1, the youngest, needed shorter sessions. Elephant 5, the adult, struggled with distractions.

These differences highlight how individual predispositions shape learning outcomes.

But nurture matters too. By consistently rewarding desired actions, the researchers taught the elephants a complex sequence of behaviours.

This success shows that environmental factors and learning experience can reshape behaviour, even in animals with distinct innate characteristics.

Individual and situational explanations

The study’s design and findings also relate to the individual vs. situational debate. That debate asks whether behaviour is driven mainly by internal factors, like personality and disposition, or by external factors, like environmental cues and social context.

Both sides have a point here.

The study’s focus on individual differences in learning speed lends support to the individual side. The variations observed suggest that traits like temperament or cognitive ability shape how an animal responds to training.

Context matters too.

The study also acknowledges situational influence: the researchers note real distractions during training, such as tourists or other animals nearby. These external factors could affect an elephant’s focus and performance, regardless of its individual predispositions.

Reductionism versus holism

Fagen et al. (2014) primarily employs a reductionist approach by focusing on a specific behavioral mechanism, Secondary positive reinforcement (SPR), to explain and modify elephant behavior.

This focus on a single learning principle exemplifies a reductionist approach, simplifying the explanation of behavior to a specific mechanism without fully considering the potential influence of other factors.

Reductionism has real limits.

However, the study also demonstrates some aspects of a holistic perspective by recognizing the role of individual differences and situational factors in the training process.

The researchers acknowledge the individual differences in learning speed and performance among the elephants, suggesting that factors like personality, cognitive abilities, and previous experiences could influence their responses to training.

That is the classic trade-off in behaviourist explanation: a clean, testable mechanism, at the cost of saying much about what the elephant experiences.

Free Will Versus Determinism

Operant training is a deterministic technology: behaviour is brought under the control of its consequences, not free choice in any philosophical sense.

Yet the study’s own framing is voluntary participation, with each elephant free to decline simply by not offering the behaviour.

That is the paradox. That tension is exactly why the study reads as a welfare advance, not merely an efficient form of control.

A strict determinist would call the elephant’s cooperation nothing more than a trained response, with no meaningful sense in which it is freely choosing anything at all.

Both readings describe the same behaviour.

Fagen et al. (2014) do not try to resolve this philosophical question. They simply note that the elephants could, in practice, walk away, and treat that practical freedom as the ethically important fact.

Ethics and Anthropomorphism

Using animals in training research always raises questions of welfare, cost and benefit, and the animal’s right to withdraw.

Here the balance is unusually favourable, because the whole intervention is designed to reduce coercion rather than increase it.

That is a genuine ethical strength.

A residual caution is anthropomorphism: reading the elephants’ cooperation as willing “consent” risks over-interpreting behaviour that is, mechanistically, just reinforced responding to a whistle and a banana.

An elephant that walks away is expressing a trained option, not necessarily a considered choice in the human sense of the word.

The two readings are not mutually exclusive.

The safest conclusion treats the welfare benefit as real, less coercion, less fear, without claiming to know what the experience is like for the elephant itself.

Keep Learning

To help reinforce your understanding and prepare for potential exam questions, here are some practice questions related to this study for AICE Psychology:

1. From the Fagen study, what type of conditioning was investigated? (1 mark)
2. How were session times measured in the Fagen study? (2 marks)
3. How many elephants were in the study? (1 mark)
4. What specific behaviors were the elephants trained to perform? (3 marks)
5. What were the primary and secondary reinforcers used in the study? (4 marks)
6. Outline what quantitative data were collected in the Fagen et al. (2014) study. (3 marks)
7. Outline one strength and one weakness of the research method used in the Fagen et al. (2014) study. (10 marks)
8. Explain how secondary positive reinforcement was used in the Fagen et al. (2014) study. (4 marks)
9. Identify two features of the sample used in the Fagen et al. (2014) study. (2 marks)

Remember to use specific details from the study when answering these questions. Practice structuring your responses clearly, using psychological terminology correctly, and balancing description with evaluation where required.

References

Ayllon, T., & Azrin, N. H. (1968). The token economy: A motivational system for therapy and rehabilitation. Appleton-Century-Crofts.

Fagen, A., Acharya, N., & Kaufman, G. E. (2014). Positive reinforcement training for a trunk wash in Nepal’s working elephants: Demonstrating alternatives to traditional elephant training techniques. Journal of Applied Animal Welfare Science, 17(2), 83-97. https://doi.org/10.1080/10888705.2014.856258

Skinner, B. F. (1938). The behavior of organisms: An experimental analysis. Appleton-Century-Crofts.

Skinner, B. F. (1953). Science and human behavior. Macmillan.

Thorndike, E. L. (1911). Animal intelligence: Experimental studies. Macmillan.

Wilson, M. L., Perdue, B. M., Bloomsmith, M. A., & Maple, T. L. (2015). Rates of reinforcement and measures of compliance in free and protected contact elephant management systems. Zoo Biology, 34(5), 431-437. https://doi.org/10.1002/zoo.21229

Wilson, M. L., Kelling, A. S., & Perdue, B. M. (2026). A longitudinal exploration of protected contact and positive reinforcement training in elephants. Zoo Biology, 45(2), 203-207. https://doi.org/10.1002/zoo.70050

Study Summary Sheet

Original Journal Article