Approach–Avoidance Conflict in Dogs: Competing Motivations

1. What Approach–Avoidance Conflict Is

1.1 The Definition

Approach–avoidance conflict occurs when a single stimulus carries both attractive and aversive value, so that the same object generates competing motivations simultaneously. The dog is not undecided in the sense of lacking information; it is being pulled in two directions by the same thing.

Everyday examples are constant: the impulse to investigate a stranger against fear of that stranger; the motivation to reach another dog against restraint on the lead; the pull of food against caution about the person standing over it (much of what presents as reactivity has this structure).

1.2 Why It Destabilizes Behavior

An animal can approach or withdraw. It cannot do both. When the two motivations are close in strength, behavior becomes unstable rather than intermediate — which is why conflict produces oscillation, interruption, and abrupt switching rather than a smooth compromise (and why arousal often rises alongside it).

1.3 Conflict Is Not the Same as Ambivalence

The everyday word for a dog that cannot decide is confused, and confusion implies a lack of information. Conflict is the opposite situation: the dog has information about two things it wants, and they point in different directions.

That distinction changes what to do. More information does not resolve a conflict, and reducing one of the two motivations does (because transfer across contexts is not automatic). Showing a frightened dog more of the thing it is frightened of is the commonest version of that error, and it is usually offered in good faith as reassurance.

1.4 Why It Is Not a Diagnosis

Conflict describes a situation the animal is in rather than a condition it has. The same dog is in conflict at the vet and not in conflict at home, and calling a dog conflicted as though it were a trait misdescribes the phenomenon.

What can be a stable property is how readily a given dog enters conflict and how it behaves once there, which is a temperament question rather than a description of the state (with individual variation doing more work than group means suggest). Keeping the two apart matters in a consultation, because one is changeable by arranging the situation and the other is not.

2. Lewin's Framework and What Has Actually Been Tested

2.1 The Three Conflict Types

Kurt Lewin's field theory (1931) distinguished three forms. Approach–approach: two desirable options, the easiest to resolve. Avoidance–avoidance: two undesirable options, harder. Approach–avoidance: one object with both qualities, and the form most relevant to problem behavior in dogs.

Lewin's account posits opposing gradients whose strength changes with proximity, with the avoidance gradient typically rising more steeply as the animal nears the goal. That asymmetry predicts something owners see constantly: a dog that appears composed at distance and deteriorates as it closes.

2.2 What the Recent Test Actually Showed

The force-field mechanism was long asserted and rarely tested. Enisman, Levy and Kleiman (2024) provided the first direct empirical test using mouse-tracking across five studies with 534 human participants — and it is worth being precise about what they tested, because this study is frequently cited for more than it shows. They compared approach–approach against avoidance–avoidance conflicts, and found the predicted signatures of the force-field mechanism, including oscillations and returns toward the midpoint between options, in the avoidance–avoidance case.

That is a genuine validation of Lewin's mechanism, in humans, for a different conflict type than the one this article concerns. The oscillation seen in a conflicted dog is described behaviorally and fits the same theoretical picture; it has not been validated by this study or, as far as canine research goes, by any equivalent one.

2.3 Why a 1931 Framework Is Still Being Used

The three-type scheme has lasted because it is descriptive rather than explanatory: it sorts situations by the sign of the competing motivations without claiming a mechanism. Schemes of that kind age well.

It also means the framework cannot be wrong in any interesting sense, which is a limitation rather than a strength, however useful the scheme is for organizing cases. A classification that no observation could contradict does not carry evidential weight. Its value is as a shared vocabulary for describing situations, and that is a real service as long as nobody mistakes it for a finding.

2.4 What the Recent Test Adds

Testing the framework with human hand movement trajectories (Enisman, Levy & Kleiman, 2024) supplies something the original did not: a continuous record of the decision as it unfolds rather than only its outcome.

The canine equivalent would be recording a dog's approach path in fine detail — speed, hesitations, deviations — rather than scoring whether it approached. Nobody has done it, and it is technically straightforward. Video and simple tracking software would suffice, and the result would be the first quantitative description of canine conflict rather than a categorical one.

3. Neural Substrates: What Rodent Work Suggests

The circuitry of approach–avoidance conflict has been mapped almost entirely in rodents. What follows is a framework rather than a description of the dog brain, and the distinction matters more here than in most topics because the popular presentation routinely omits it.

3.1 Medial Prefrontal Cortex

This mechanism is described in rodents; a comparable role in dogs is assumed but not demonstrated. Rodent work implicates the prelimbic region in risky decision-making under conflict, with individual animals falling into distinct phenotypes — some continuing to seek food under threat, others suppressing it entirely — and manipulation of prelimbic activity shifting behavior in the expected directions (Fernandez-Leon et al., 2021). The infralimbic region is associated with inhibiting defensive responses and retrieving extinction memories.

Nothing comparable has been measured in dogs. The general principle that frontal function degrades under uncontrollable stress is well established across rodents, primates, and humans (Arnsten, 2009) and is applied to dogs by extension (with the homology caveats set out separately).

3.2 Context Versus Object

Optogenetic work in rodents suggests a division of labour: ventral hippocampus more involved where the conflicting values attach to a context, perirhinal cortex where they attach to an object (Dhawan et al., 2023). If this transfers, it would predict that a dog conflicted about a specific person and a dog conflicted about a specific room are running partly different processes — a testable idea, untested in dogs.

3.3 Amygdala and Striatum

Lesion work in a foraging paradigm implicates the central amygdala in conflict-specific behaviors such as hesitant diagonal approach and stretched posture, distinct from approach and avoidance themselves (Kimm et al., 2023). Studies tracking conflict strategies over extended training report stable individual differences, with more behaviorally flexible animals showing greater ventral striatal and midline thalamic involvement (Bravo-Rivera et al., 2021).

The consistent finding worth carrying over is the individual variation: under identical conflict, animals resolve it differently and stably. That matches what trainers observe (and what temperament research predicts).

3.4 What the Rodent Work Predicts for Dogs

If the circuitry is conserved, three things should follow in dogs: conflict behavior should worsen under stress as prefrontal regulation is impaired, it should be sensitive to context rather than to the object alone, and it should be modifiable through the same associative processes as fear itself.

All three are consistent with clinical observation in dogs, and none has been tested against an alternative account in this species (where chronic load is treated at length). Consistency with observation is weak support, since a competing account that also predicted all three would look identical from here.

3.5 Why Context Versus Object Matters Practically

The distinction that emerges most clearly from the rodent work is that conflict is bound to situations more than to things. A dog is not in conflict about a person; it is in conflict about being approached by that person in that place with that much room to move. Remove the corridor and the same approach may produce nothing.

That predicts what handlers repeatedly find: changing the setting resolves a conflict that changing the dog's relationship with the object does not (because transfer across contexts is not automatic). It also explains why work done in one location so often has to be repeated in the next, which households experience as the training not sticking.

4. What Conflict Looks Like in Dogs

An important caution first: not every ambiguous behavior reflects motivational conflict. Some arises from incomplete information, from competing learned responses, or from a state that is simply changing quickly. The behaviors below are associated with conflict; none is diagnostic alone.

4.1 Approach–Withdrawal Oscillation

Movement toward the stimulus, then away, then toward again. This is one of the most direct behavioral expressions of competing gradients and is most visible in hesitant greetings.

4.2 Freezing and Stretched Posture

Freezing — complete cessation of movement apart from respiration — occurs where the competing motivations approximately cancel. Stretched posture, in which the body stays back while the head extends forward, is a compromise: the dog investigates while remaining configured to retreat.

4.3 Displacement Behavior

Normal behavior performed out of context: sniffing the ground before a greeting, scratching, yawning, lip licking, brief grooming. These are classic correlates of conflict and their interpretation requires restraint. Sniffing may be a way of declining social pressure rather than a stress signal, and a single displacement behavior does not establish an emotional state (the single-cue problem in general).

4.4 Concurrent Approach and Avoidance Signals

The dog orients toward the stimulus while simultaneously showing flattened ears, sclera visible, lowered body, and tucked tail. Because the dog is facing the trigger, this is frequently read as friendly. It indicates high conflict (a clear case of behavior and emotional state coming apart).

The description should stay observational. Terms implying deliberation — that the dog is being "intentionally" ambivalent — import a claim about the dog's reasoning that the behavior does not support.

4.5 Redirection

With approach and avoidance balanced, behavior may be directed at a neutral target instead: sudden self-scratching, barking at nothing, snapping at the lead. Redirection is well recognized clinically under restraint and high arousal (and connects directly to frustration mechanisms).

4.6 Suppression

Where conflict is chronic and escape unavailable, some dogs stop displaying it. They appear settled while the underlying state persists. This is masking rather than resolution, and the modern account of learned helplessness applies where the unavailability of control is the defining feature (Maier & Seligman, 2016) (the revised account in full).

The behavioral markers used to code this in training-methods research — turning the body or head away, stepping away, crouching, rolling onto the side or back, and slow paw lifts, each combined with lowered posture, ears back, and tail down — come from ethograms developed for welfare assessment (Vieira de Castro et al., 2020).

4.7 Why Suppression Is the Dangerous One

Of the behaviors listed, suppression is the one that looks like resolution and is not. A dog that has stopped signalling conflict may have resolved it, may have learned that signalling is punished, or may have shut down.

The three are indistinguishable in the moment and diverge later — typically when the situation changes slightly or when the suppressing contingency is absent — which is why a dog that has become quiet under a training regime deserves more scrutiny than one that is still oscillating visibly (where arousal is treated as its own variable). A dog still showing conflict is at least still communicating, and that is worth protecting rather than removing.

4.8 Reading the Sequence, Not the Snapshot

Each sign in this chapter is a moment. Conflict is a pattern across moments, and a photograph of a dog with a stretched posture establishes almost nothing.

What identifies conflict is the alternation: forward, stop, back, forward again. Anyone assessing from still images or brief clips is working with the wrong unit of observation, and the body-language infographics that circulate widely are built entirely from single moments.

5. Typical Conflict Situations

5.1 Where It Occurs

Lead reactivity — Approach: the pull to reach the other dog. Avoidance: fear of it, discomfort from lead pressure, or frustration at restraint. Signs: lunging and barking with ears back and tail low at the same time — a state problem, not a status problem (as the dominance literature makes clear).

Resource guarding with ambivalence — Approach: the dog wants the resource. Avoidance: it is wary of the person nearby. Signs: eating with sclera showing, tense body, stopping if the person shifts.

Hesitant greeting — Approach: interest in the person. Avoidance: the person's posture, direct gaze, or reaching hand. Signs: stretched approach, low tail, ears back, then retreat or displacement.

Barrier frustration — Approach: the stimulus on the far side. Avoidance: the restraint itself, plus whatever the dog anticipates on contact. Signs: fence-running, whining, redirection onto the barrier.

Handling and veterinary procedures — Approach: a learned history of tolerating it. Avoidance: fear or pain. Signs: lying on the side with tense musculature, paw lift, sclera visible, a growl that is easy to punish and expensive to lose (and pain deserves ruling out first).

5.2 The Common Thread

The situations listed above share a structure: something the dog wants or must accept is co-located with something it wants to avoid. Veterinary handling, a resource near a person, a greeting with an ambivalent dog, a doorway with something startling beyond it.

Recognizing the structure is more useful than memorizing the list, because the list is not exhaustive and every household produces its own variants (which is where cumulative load becomes visible). Once the structure is visible, a household can usually name three or four situations nobody asked about, and those are frequently the ones that matter most.

5.3 Medical Situations Are the Frequent Case

Veterinary and grooming contexts generate more conflict than almost anything else, because the dog is restrained, handled by a stranger, in an unfamiliar place, and often already uncomfortable.

Where a dog's conflict behavior is concentrated in those settings, pain and previous experience deserve investigation before anything is attributed to temperament (where pain turns out to be a recurring differential). A dog that is fine everywhere except on a table is telling you something about tables, or about what has happened on them, and neither answer is a statement about the dog's character.

6. Assessing Conflict

6.1 What to Look For

Standardized questionnaires. The C-BARQ (Hsu & Serpell, 2003) provides scores across behavior categories including stranger-directed fear and stranger-directed aggression — constructs that are, in effect, conflict descriptions. It is validated and widely used, with the usual limitation that it captures owner report rather than direct observation.

Behavioral testing. Souza et al. (2023) analyzed video of 70 pet dogs exploring freely for 30 seconds while an actor expressed a positive, neutral, or negative emotion, and found behavior differed systematically by valence — with tail elevation and physical contact during sniffing emerging as candidate markers of a positive state. This is consistent with dogs' approach-related behavior tracking the emotional valence of the situation. It is not itself a conflict paradigm, and no validated canine approach–avoidance conflict test currently exists.

The Strange Situation Procedure. Introducing a stranger and separating the dog from the owner creates mild, controlled conflict between exploring and staying near the attachment figure. Reunion behavior is the informative part (the attachment framework in detail).

Physiological measures. Heart rate is the most practical and, per Flint et al. (2024), the arousal indicator that holds across positive and negative valence — which matters here, since conflict involves both. Cortisol and the common behavioral stress signs in that study differentiated arousal only under negative valence.

6.2 What an Assessment Cannot Tell You

Observing conflict identifies the state and not its content. Which two motivations are competing, how strong each is, and what history produced them are not readable from the behavior.

Those come from the history and from testing what changes the behavior, which is why an assessment without a conversation is incomplete (which is what operationalizing a behavioral claim is for). A video sent by an owner shows the state and none of the content, which is worth saying to anyone offering remote assessments on that basis. What it can establish is whether conflict is present at all, and that is not nothing.

6.3 Individual Differences Are Large

Dogs differ in how much competing motivation they tolerate before behavior destabilizes, and in what they do when it does. One dog freezes, another oscillates, a third redirects.

Comparing a dog against another dog is therefore weak; comparing it against itself in a less demanding version of the same situation is strong (with individual variation doing more work than group means suggest). That within-animal comparison is the same principle that governs assessment throughout this field and it is repeatedly abandoned in favour of breed expectations.

7. How Conflict Is Measured

7.1 The Inference Problem

Conflict is a state of competing motivations, and no measure reads it directly. What an observer records is behavior that is compatible with conflict — oscillation, freezing, displacement activity — and the inference from one to the other is where the difficulty sits (which is what operationalizing a behavioral claim is for).

7.2 Why Displacement Behavior Is Ambiguous

Scratching, yawning, sniffing the ground and lip-licking are the classic indicators, and each of them also occurs for ordinary reasons. A dog scratches because it itches, yawns because it is tired, and sniffs because something smells interesting.

What makes these behaviors informative is context and clustering: several of them together, out of their usual context, in a situation where two motivations plausibly compete. Any one of them alone is weak evidence. This is the reason two competent observers can watch the same dog and disagree, and it is not a failing of either of them.

7.3 What the Canine Measurement Work Has Established

Indicators of positive and negative emotional state have been evaluated in dogs (Flint et al., 2024), and the wider positive-emotion assessment literature has been reviewed (Csoltova & Mehinagic, 2020). Validated owner questionnaires exist for temperament and behavior (Hsu & Serpell, 2003).

None of these was designed to measure conflict specifically. The field has instruments for arousal, for fear and for temperament, and none for the state this article is about (where arousal is treated as its own variable). That absence is worth registering before reading any claim about how common conflict behavior is.

7.4 Oscillation Is the Most Specific Sign

Of the behaviors described here, repeated approach and withdrawal toward the same target is the one least explicable by anything else. A dog that steps forward, stops, steps back and repeats is doing something that a single motivation does not readily produce.

It is also the sign that requires the longest observation. A snapshot catches the dog either approaching or retreating, and the pattern only appears over time (with individual variation doing more work than group means suggest). Thirty seconds of continuous watching is usually enough and is more than most assessments allow, particularly where the dog is being handled at the same time.

7.5 What Would Make Assessment Better

Two things, neither requiring equipment. Recording the sequence rather than the moment, since conflict is defined by alternation and a single observation cannot show it. And noting what changes the behavior: if increasing distance resolves it, one motivation was fear; if removing the attractive target resolves it, the other was appetitive. Neither test requires the observer to interpret a single posture.

That second test is diagnostic in a way no observation of posture can be, and it is available in any consultation. It also produces an answer the household can act on, which a description of the dog's body language does not.

8. Resolving Conflict

8.1 Recognize It Before It Escalates

Stretched posture, freezing, paw lift, sclera visible, displacement, oscillation. When these appear, the dog is not refusing — it is caught. Adding pressure at that point resolves nothing and adds a third motivation to avoid.

8.2 Change One of the Two Motivations

Conflict resolves when the balance shifts, which means acting on one side. Reduce avoidance: increase distance, remove the pressure, let the dog leave. Reduce approach: offer an alternative that is worth more and costs less. What does not work is leaving both intact and punishing the behavior that expresses them.

8.3 Desensitization and Counterconditioning

The systematic version of reducing avoidance: exposure held below the level at which conflict appears, paired with something the dog values, so that what the trigger predicts changes (the underlying fear-memory mechanics). There is controlled canine support for this approach (Butler, Sargisson & Elliffe, 2011; Shnookal et al., 2024), and Riemer (2023) found relaxation training and counterconditioning to real-life stimuli among the better-supported interventions in the noise-fear literature.

8.4 Do Not Punish Conflict Behavior

One distinction belongs here before anything else. Calmly interrupting a situation, increasing distance, or redirecting a dog to something else is not punishment – it removes the dog from the conflict rather than adding a consequence to the signalling. Punishment means applying something the dog will work to avoid, and the objection below concerns that and not management.

Punishing a growl, a freeze, or a snap suppresses the display without touching either motivation. The dog remains conflicted and has lost the signal that made its state legible. Schilder and van der Borg (2004) documented the general process in shock-trained dogs, who carried stress-related behavior into sessions where nothing happened to them (the broader fallout picture).

8.5 Test Whether It Actually Resolved

If a dog has stopped displaying conflict, distinguish resolution from masking. Stop the interaction and wait: a dog that re-engages voluntarily has changed state; a dog that stays still, disengages, or leaves has not. Behavior that has merely been suppressed tends to return when conditions permit (as the extinction literature predicts).

8.6 Manage the Environment

Distance, visual barriers, and avoiding over-threshold exposure prevent rehearsal and give the system room to settle. Management is not a failure to train; it is the precondition for training.

8.7 Watch the Human Side

Rial et al. (2025) report that dogs witnessing conflict between their caregivers showed increased stress and displacement behavior and less relaxed behavior — the dog's own approach–avoidance situation, generated by the household (consistent with what is known about emotional contagion).

8.8 Why Punishing Conflict Behavior Backfires Specifically

Punishment applied to a dog in conflict adds a third motivation to a situation that already had two too many. It does not remove either of the original ones, and it makes the situation less predictable for the dog rather than more.

What it reliably does is suppress the signalling, which removes the observer's warning while leaving the state intact. The dog still wants both things and now has a reason to hide that it does. That can contribute to bites that appear to come from nowhere. Where it does, nothing came from nowhere; the sequence that would have announced it was trained out earlier, often by someone acting on advice. Other routes to an unwarned bite exist – pain, a startled dog, a signal nobody present could read – and suppression is one contributor among them.

8.9 What Resolution Actually Looks Like

A resolved conflict shows as smooth, committed movement in one direction — approach without hesitation or a clean decision to leave. Not as a dog that approaches while still displaying displacement behavior. That dog has been persuaded to override the conflict rather than relieved of it, and the override is not durable.

Testing for that rather than for the absence of obvious distress is the difference between a plan that worked and one that produced compliance (which is where cumulative load becomes visible). The test is also cheap: present a slightly easier version of the same situation and watch whether the approach is clean.

9. Which Findings Come From Which Species

9.1 The Split in This Article

The behavioral picture in this article is canine. The framework is human, and the neural account is almost entirely rodent. Keeping the three apart is what allows the practical chapters to stand on their own.

This is an unusually clean example of a pattern that runs through the collection: the part that explains is borrowed, and the part that is measured in dogs is descriptive.

9.2 The Framework Is Human and Old

The three conflict types that organize this topic come from human developmental psychology of the 1930s (Lewin, 1931), and the recent test of the framework used human participants and hand movement trajectories (Enisman, Levy & Kleiman, 2024).

It is a useful organizing scheme with almost a century of currency and it was never validated in dogs. That is not an argument against using it; it is an argument for not treating its categories as measured entities. A dog does not have an approach-avoidance conflict in the way it has a temperature.

9.3 The Neural Account Is Rodent

Medial prefrontal involvement in approach-avoidance decisions, the role of perirhinal cortex, central amygdala modulation of conflict-like behavior and the threat-related circuitry all come from rodent work (Bravo-Rivera et al.; Dhawan et al., 2023; Kimm, Kim & Choi, 2023; Fernandez-Leon et al.), with the prefrontal impairment account drawn from rodent and human research (Arnsten, 2009).

Five of the article's sources are rodent studies. No canine study has imaged or recorded from any of these structures during an approach-avoidance situation, and none is likely to soon: the situation is defined by the animal moving freely, which is precisely what imaging does not permit.

9.4 What Was Measured in Dogs

Behavior in ambiguous and conflicting situations has been examined in dogs (Souza et al., 2023; Rial, Damián & Bentosela, 2025), emotional-state indicators have been evaluated (Flint et al., 2024; Csoltova & Mehinagic, 2020), and the intervention literature is canine throughout (Butler, Sargisson & Elliffe, 2011; Shnookal et al., 2024; Riemer, 2023).

The training evidence is likewise canine (Vieira de Castro et al., 2020; Schilder & van der Borg, 2004), which matters because the practical chapter rests on it rather than on the neuroscience.

9.5 Why the Split Protects the Argument

The claims a practitioner needs — that conflict behavior is a state rather than disobedience, that punishing it removes the warning rather than the state, that changing one of the two motivations is the route out — sit in the canine column.

They survive whatever happens to the rodent circuitry, and leading with the circuitry would put a solid recommendation on a borrowed footing (which is what operationalizing a behavioral claim is for). The neuroscience is worth including because it explains why the behavioral picture looks as it does, not because the recommendations depend on it.

10. Conflict at a Glance

The phenomenon — One stimulus with both attractive and aversive value, generating incompatible motivations simultaneously. Not indecision, not inconsistency, not disobedience.

Behavioral markers — Approach–withdrawal oscillation; freezing; stretched posture; displacement behaviors; concurrent approach and avoidance signals; redirection; and, where conflict is chronic and inescapable, suppression that resembles calm.

Evidence status — Theory from human psychology (Lewin, 1931), with the force-field mechanism first tested in humans for a different conflict type (Enisman et al., 2024). Circuit-level work is rodent (Fernandez-Leon et al., 2021; Dhawan et al., 2023; Kimm et al., 2023; Bravo-Rivera et al., 2021). Canine material is behavioral, and no validated canine conflict paradigm exists.

Resolution — Shift one motivation rather than suppressing the display: increase distance or reduce threat to lower avoidance, or offer a better and cheaper alternative to lower approach. Counterconditioning and desensitization are the systematic version.

What not to do — Punish the conflict behavior. It removes the signal, leaves both motivations intact, and raises the probability of escalation without warning.

11. Research Gaps and Critical Appraisal

The neural account is entirely rodent. Prefrontal, hippocampal, perirhinal, amygdalar, and striatal contributions to conflict come from rodent studies. No canine equivalent exists, and the extension rests on conserved mammalian architecture rather than on canine data.

The framework's key test was in humans and on a different conflict type. Enisman et al. (2024) validated Lewin's force-field mechanism for approach–approach versus avoidance–avoidance conflicts using mouse-tracking. It is regularly cited as though it demonstrated approach–avoidance oscillation; it did not.

There is no validated canine conflict paradigm. Souza et al. (2023) shows behavior tracks emotional valence in dogs; it is not a conflict test. Assessment currently relies on owner-report instruments and ethogram coding.

Displacement behaviors are ambiguous by construction. They are normal behaviors out of context, and their function is often unclear. Treating a yawn or a scratch as evidence of a specific state overreads them.

Positive states remain hard to measure. No validated indicator of positive emotion in dogs exists (Csoltova & Mehinagic, 2020), and the standard stress indicators differentiated arousal only under negative valence (Flint et al., 2024) — which is a particular problem in conflict, where both valences are present at once.

Cognitive dissonance in dogs is unresolved. Whether dogs experience anything like dissonance is debated, and effects sometimes attributed to it may reflect simpler contrast mechanisms. Nothing in the practical recommendations depends on the answer.

Individual variation is stable and large. Rodent work finds consistent conflict-resolution phenotypes, and the same is evident clinically — which limits how far any population-level description maps onto an individual dog (as behavioral flexibility research also suggests).

No instrument measures conflict in dogs. Validated canine tools exist for temperament (Hsu & Serpell, 2003) and for emotional-state indicators (Flint et al., 2024), and none targets competing motivations specifically.

The framework has never been tested in this species. The three-type scheme is human and nearly a century old (Lewin, 1931), and its recent empirical test used human participants (Enisman et al., 2024).

The neural account is rodent throughout. Five of this article's sources are rodent studies, and no canine work has recorded from the structures involved during an approach-avoidance situation.

Displacement behaviors are not specific. Each occurs for ordinary reasons as well, and no study has established how reliably clusters of them indicate conflict rather than general arousal.

12. Conclusion

Approach–avoidance conflict is not a disorder or a training failure. It is what happens whenever something a dog wants is also something it has reason to avoid, which in a human household is often. The behavior it produces — hesitating, freezing, approaching in a stretched crouch, sniffing the ground, snapping at the lead — reads as contradictory only if you assume a single motivation must be driving it. Read as competition between two, it becomes legible and, more usefully, actionable: conflict resolves when one side of the balance changes, and adding pressure changes neither. The evidence base deserves the same clarity. The theory is human, its central mechanism was tested in humans on a different conflict type, the circuitry is rodent, and the canine contribution is behavioral observation without a validated paradigm. That is a reasonable foundation for the practical conclusions, which do not depend on the neuroscience: increase distance, change what the trigger predicts, give the dog somewhere to go, rule out pain, and treat a dog that has gone quiet as a question rather than an answer.

Key Insights (Takeaways)

• Conflict behavior is competition between two motivations, not indecision or defiance. Oscillation, freezing, stretched posture, and displacement all follow from a dog being pulled toward and away from the same thing — and a dog facing a trigger with flattened ears and visible sclera is in high conflict, not being friendly.

• The evidence is thinner than the confident presentation suggests. Lewin's framework is human, its force-field mechanism was first tested in humans and for approach–approach versus avoidance–avoidance conflicts specifically (Enisman et al., 2024), the circuit work is rodent, and no validated canine conflict paradigm exists.

• Resolution means shifting one motivation. Reduce avoidance through distance and counterconditioning, or reduce approach by offering something better and cheaper. Leaving both in place and punishing the resulting behavior addresses neither.

• Displacement behaviors are correlates, not diagnoses. Sniffing, yawning, and scratching are normal behaviors out of context; a dog sniffing the ground may be declining social pressure rather than signaling distress, and no single cue establishes a state.

• A dog that stops showing conflict has not necessarily stopped being conflicted. Test it: stop the interaction and see whether the dog re-engages. Suppressed conflict costs the warning signals that made the dog's state readable, and household tension is itself a source of it (Rial et al., 2025).

References

Arnsten, A. F. T. (2009). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience, 10(6), 410–422. https://doi.org/10.1038/nrn2648

Bravo-Rivera, H., Rubio Arzola, P., Caban-Murillo, A., Vélez-Avilés, A. N., Ayala-Rosario, S. N., & Quirk, G. J. (2021). Characterizing different strategies for resolving approach-avoidance conflict. Frontiers in Neuroscience, 15, 608922. https://doi.org/10.3389/fnins.2021.608922

Butler, R., Sargisson, R. J., & Elliffe, D. (2011). The efficacy of systematic desensitization for treating the separation-related problem behaviour of domestic dogs. Applied Animal Behaviour Science, 129(2–4), 136–145. https://doi.org/10.1016/j.applanim.2010.11.001

Csoltova, E., & Mehinagic, E. (2020). Where do we stand in the domestic dog (Canis familiaris) positive-emotion assessment: A state-of-the-art review and future directions. Frontiers in Psychology, 11, 2131. https://doi.org/10.3389/fpsyg.2020.02131

Dhawan, S. S., Pinter, C., Lee, A. C. H., & Ito, R. (2023). Male rodent perirhinal cortex, but not ventral hippocampus, inhibition induces approach bias under object-based approach-avoidance conflict. eLife, 12, e81467. https://doi.org/10.7554/eLife.81467

Enisman, M., Levy, A., & Kleiman, T. (2024). Hand movement trajectories illustrate the mechanism underlying Kurt Lewin's distinction between approach–approach and avoidance–avoidance motivational conflicts. Journal of Personality and Social Psychology, 127(2), 239–258. https://doi.org/10.1037/pspa0000392

Fernandez-Leon, J. A., Engelke, D. S., Aquino-Miranda, G., Goodson, A., Rasheed, M. N., & Do Monte, F. H. (2021). Neural correlates and determinants of approach–avoidance conflict in the prelimbic prefrontal cortex. eLife, 10, e74950. https://doi.org/10.7554/eLife.74950

Flint, H. E., Weller, J. E., Parry-Howells, N., Ellerby, Z. W., McKay, S. L., & King, T. (2024). Evaluation of indicators of acute emotional states in dogs. Scientific Reports, 14(1), 6406. https://doi.org/10.1038/s41598-024-56859-9

Hsu, Y., & Serpell, J. A. (2003). Development and validation of a questionnaire for measuring behavior and temperament traits in pet dogs. Journal of the American Veterinary Medical Association, 223(9), 1293–1300. https://doi.org/10.2460/javma.2003.223.1293

Kimm, S., Kim, J. J., & Choi, J.-S. (2023). The central amygdala modulates distinctive conflict-like behaviors in a naturalistic foraging task. Frontiers in Behavioral Neuroscience, 17, 1212884. https://doi.org/10.3389/fnbeh.2023.1212884

Lewin, K. (1931). Environmental forces in child behavior and development. In C. Murchison (Ed.), A handbook of child psychology (pp. 590–625). Clark University Press.

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

Rial, L. A., Damián, J. P., & Bentosela, M. (2025). Affiliative and stress behaviors of domestic dogs when they witness a human conflict between their caregivers. Applied Animal Behaviour Science, 289, 106680. https://doi.org/10.1016/j.applanim.2025.106680

Riemer, S. (2023). Therapy and prevention of noise fears in dogs — A review of the current evidence for practitioners. Animals, 13(23), 3664. https://doi.org/10.3390/ani13233664

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

Shnookal, J., Tepper, D., Howell, T., & Bennett, P. (2024). Counterconditioning-based interventions for companion dog behavioural modification: A systematic review. Applied Animal Behaviour Science, 276, 106305. https://doi.org/10.1016/j.applanim.2024.106305

Souza, P., Guo, K., Mills, D. S., Resende, B., & Albuquerque, N. (2023). How do dogs behave when presented with situations of different emotional valences? Animals, 13(6), 1027. https://doi.org/10.3390/ani13061027

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), e0225023. https://doi.org/10.1371/journal.pone.0225023