Conflict and Reconciliation in Dogs: How Dogs Resolve Disputes

1. Why Conflict Management Matters in Social Animals

1.1 Conflict Is Part of Group Life

Animals that live in stable groups compete for food, resting places, mates and attention, and some of that competition turns into aggression. A widely cited perspective paper reframed aggression as a tool of competition and negotiation: when survival depends on mutual assistance, aggression is constrained by the need to maintain beneficial relationships (de Waal, 2000). According to the framework by Aureli and colleagues, as applied in canid studies, conflicts carry costs: a risk of injury, increased stress and damage to relationships (Aureli et al., 2002). Conflict management is the umbrella term for behavior that reduces those costs. It includes strategies before a conflict escalates, such as signals, submission and moving away, and strategies after it, such as friendly contact between former opponents, contact with uninvolved group members or simply keeping apart.

1.2 Where the Term Reconciliation Comes From

The term reconciliation was coined in chimpanzee research for friendly contact between former opponents soon after a conflict, contact that may help to repair a relationship damaged by the encounter (de Waal & van Roosmalen, 1979, as summarized by Cafazzo et al., 2018). The word carries an interpretation: humans reconcile to restore a relationship, and the label invites the assumption that animals do the same. Lazzaroni et al. (2017) therefore recommended the neutral expression "post-conflict opponent affiliation" wherever a function has not been shown. This article uses "reconciliation" where the studies used it and keeps the question of function open. That matters for dogs in particular, because popular accounts describe dogs that apologize or make up with a lick, assuming intentions that no study has measured (why behavior is not a direct readout of emotion).

1.3 Three Meanings of Conflict

"Conflict" is used in at least three senses. This article is mainly about agonistic conflict: an aggressive interaction, with or without physical contact, between two animals. A second sense is motivational conflict within one dog that is drawn toward and away from something at the same time (what motivational conflict means). A third is conflict between a dog and a person. Findings from these areas cannot be swapped, so conflicts involving people are treated separately in Section 6.

2. How Post-Conflict Behavior Is Measured

2.1 The Post-Conflict and Matched-Control Comparison

Friendly contact after a fight is not, on its own, evidence of reconciliation. Animals that live together interact often anyway. The standard method therefore uses a comparison. After an aggressive interaction, the victim, and in some studies both opponents, is observed for a fixed post-conflict period (PC), in canid studies usually 10 minutes. A matched-control observation (MC) of the same animal follows on another day, at a comparable time and without a preceding conflict. The method goes back to work on rhesus macaques (de Waal & Yoshihara, 1983) and was later refined (Veenema et al., 1994). Details differ: at the Wolf Science Center, the MC took place on the next possible day at the same time, only if there had been no agonistic interaction in the 10 minutes before (Cafazzo et al., 2018); in a zoo wolf study, the rule was no agonistic interaction in the preceding 30 minutes (Cordoni & Palagi, 2008).

2.2 Attracted, Dispersed and the Corrected Conciliatory Tendency

Each PC–MC pair is classified. It is "attracted" if the former opponents made friendly contact earlier in the PC than in the MC, or only in the PC; "dispersed" if contact came earlier in the MC, or only in the MC; and "neutral" otherwise. Reconciliation is shown when attracted pairs clearly outnumber dispersed ones, not merely because friendly contact occurs after fights.

To compare groups and species, researchers use the corrected conciliatory tendency (CCT): attracted minus dispersed pairs, divided by all pairs. Veenema et al. (1994) proposed a corrected measure because, unlike the original one, it is independent of baseline affiliation and therefore suitable for comparisons. A CCT of 50% means that the proportion of attracted pairs exceeds that of dispersed pairs by 50 percentage points. It does not mean that reconciliation occurred in exactly half of all conflicts. A CCT near zero means that attracted and dispersed pairs roughly balance; friendly post-conflict contact may still occur.

2.3 Pitfalls: Definitions, Time Windows and Context

Results depend on choices that are easy to overlook. The first is what counts as aggression and as friendly contact. If directed barking is coded as light aggression and anogenital sniffing as affiliation, more interactions qualify as conflicts and more contacts as friendly. Both choices were made in one dog study, and the authors of a later study suggested that they may have inflated both conflicts and reconciliations (Cafazzo et al., 2018, discussing Cools et al., 2008).

The second is the time window. Veenema et al. (1994) found that two of three methods they compared may fail with long observation periods and recommended attention to timing. In canids, friendly reunions happen mostly within the first one or two minutes (Section 3.5), so a long window that is not analyzed minute by minute can dilute a real effect. The third is context: conflicts over food and conflicts in other situations can give very different results (Section 3.6), and pooling them hides the difference. Finally, the method needs many conflicts. A group that rarely fights produces few PC–MC pairs, as the dog data in Section 4 show.

2.4 Measuring Stress After Conflict

In long-tailed macaques, victims increased tension-related activities such as scratching, body shaking and self-grooming after conflicts and had a higher risk of renewed attacks. Reconciliation and redirection, aggression turned against an uninvolved animal, reduced these rates and the likelihood of renewed attacks, with reconciliation the more effective (Aureli & van Schaik, 1991). This is the origin of self-directed behavior as a post-conflict anxiety measure. Transfer to canids is not straightforward. In Arctic wolves, scratching made up only 0.74% of stress-related behaviors in post-conflict periods, while lip licking made up 75.82% (Lazzaroni et al., 2017). Which behaviors indicate stress in a canid after a conflict is a measurement question of its own (what behavioral and physiological stress measures can show). Statements about post-conflict anxiety in dogs have no direct study behind them.

3. Reconciliation in Wolves

Four groups of studies applied the PC–MC method to wolf packs, and reconciliation was found in all of them.

3.1 A Zoo Pack in Pistoia

The first canid study followed a captive pack at Pistoia Zoo for 13 months, from March 2005 to May 2006 (Cordoni & Palagi, 2008). The pack consisted of nine adult wolves, five males and four females: captive-bred siblings and their father, a "disrupted family" without the breeding female. The researchers recorded 3,344 conflicts and 336 PC–MC pairs. Attracted pairs outnumbered dispersed ones, and the mean CCT was 53.2%. Victims directed a larger share of their friendly behavior to the former opponent in PCs than in MCs, which indicates selective attraction. The CCT did not differ between sex combinations or between low- and high-intensity conflicts (59.2% vs 50.1%, not significant) and did not correlate with rank distance. Aggressors and victims did not differ significantly in how often they initiated contact. The limits are clear: one pack, nine related animals, one zoo.

3.2 Arctic Wolves in Olomouc

A second study observed a family pack of Arctic wolves at Olomouc Zoo for 154 hours (Lazzaroni et al., 2017). The pack had 19 wolves at the start and 14 by the end; one breeding female was removed before data collection after being badly injured by the other, and an adult male was removed because of continuous mobbing. The researchers recorded 585 aggressions with physical contact and 386 without and collected 133 PC–MC pairs. The mean CCT was 46.87%. Opponent affiliation occurred mainly outside feeding, more often when the victim was dominant and the aggressor subordinate, and was more often initiated by the victim. The authors said their results should be viewed "with a measure of caution": one zoo pack, a changing composition and no replication in the wild.

3.3 Two Wild Packs in Yellowstone

In two free-ranging Yellowstone packs, the Druid Peak pack with 16 wolves and the Blacktail Deer Plateau pack, researchers obtained 34 PC and 34 MC periods from 106 hours of video in one winter (Baan et al., 2014). Of all conflicts in PC periods (34 initial, 7 renewed), 61.0% (25) were followed by reconciliation, 4.9% (2) by consolation and 19.5% (8) by solicited consolation, in which the victim approached a bystander. The CCT was 44.1%, and reconciliation was mainly initiated by victims. The authors argued that interdependence within the pack can explain investment in conflict mitigation "without the need to invoke particular relationships of mutual value". The data come from wild animals, which is valuable, but they are few.

3.4 Unrelated Wolves at the Wolf Science Center

The fourth study observed four captive packs of unrelated, hand-raised wolves at the Wolf Science Center in Austria, with two to five animals per pack (Cafazzo et al., 2018). Among 419 aggressive interactions, there were 177 PC–MC pairs. Friendly contact between opponents occurred in 42.4% of PCs versus 19.8% of MCs, 71 of 177 conflicts were reconciled, and the mean CCT was 22.46%, ranging from −25.00% to 63.16% between individuals. Reconciliation was more likely when opponents were further apart in rank, outside feeding, and after high-intensity aggression (48 of 104 vs 22 of 71 conflicts). It was initiated by the victim in 58 cases and by the aggressor in 13. The authors attributed the lower CCT to the packs being made up of unrelated animals rather than to pack size.

3.5 Fast and Short

Friendly reunions in wolves were quick. At the Wolf Science Center, the PC–MC difference was greatest in the first minute, mean latency was 99.4 ± 145.8 seconds, and most reconciliations took place within two minutes (Cafazzo et al., 2018). In Arctic wolves, 78.56% (44 of 56) of first friendly contacts in PCs occurred within two minutes, compared with 33.33% (4 of 12) in MCs (Lazzaroni et al., 2017). In Yellowstone, the difference was significant only in minutes one to four (Baan et al., 2014). The contacts were also brief, with a mean of 8.2 ± 10.8 seconds at the Wolf Science Center. A reconciliation, if it occurs, is therefore easy to miss: not a long cuddle an hour later, but a short approach within a minute or two.

3.6 Food Changes the Picture

At the Wolf Science Center, 9 of 45 conflicts that occurred while food was available were followed by a friendly interaction, compared with 61 of 130 conflicts outside feeding (Cafazzo et al., 2018). The paper gives the food figure as a percentage that does not match these counts, so the counts are used here. In Arctic wolves, too, opponent affiliation occurred mainly outside feeding (Lazzaroni et al., 2017). One possible reading, not tested in the studies, is that a conflict over a resource that is still present is not over when the aggression stops. Either way, a pooled CCT describes neither situation well.

3.7 What Reconciliation Does in Wolves

Whether reconciliation has a function in wolves is less clear than whether it occurs. In the Arctic wolf pack, 18 cases of renewed aggression toward the victim were recorded, 6 after an opponent affiliation and 12 without, and affiliation reduced the likelihood of re-aggression (χ2 = 4.79, p = 0.03) and of redirected aggression (1 of 12 cases after an affiliation; χ2 = 8.34, p = 0.004) (Lazzaroni et al., 2017). In Yellowstone, renewed aggression did not differ between PCs with and without friendly behavior (χ2 = 0.05, P = 0.81), with only 34 conflicts (Baan et al., 2014).

Arctic wolf victims showed far more stress-related behaviors in PCs than in MCs (943 vs 360), but post-conflict affiliation did not reduce their rate (Lazzaroni et al., 2017). In the only canid study that tested it, friendly contact with the former opponent did not lower behavioral signs of stress. The stress-reducing function often attributed to reconciliation comes from primates (Aureli & van Schaik, 1991). The evidence therefore supports the statement that opponent affiliation in wolves may reduce further aggression in some settings, not that it repairs relationships or calms the animals.

4. Dogs Living in Groups: A Split Picture

4.1 The First Dog Study

Only two studies have applied the PC–MC method to conflicts between domestic dogs. The first observed three captive groups (Cools et al., 2008). According to its abstract, both reconciliation and third-party post-conflict interactions were present, familiar dogs reconciled a significantly higher proportion of conflicts than unfamiliar ones and had fewer conflicts, and most third-party affiliations involved the victim.

The full text could not be checked for this article, so its numbers are given only as reported by later authors. According to Cafazzo et al. (2018), the groups comprised 7, 7 and 6 dogs, 1,711 conflicts were recorded over two months, and 606 reconciliations were counted. Several dogs were related, and familiarity and relatedness overlapped largely. Cafazzo and colleagues recomputed mean CCTs from the paper's data at 16.6% to 29.7%, values the original authors did not report. They also noted that when a conflict was not immediately followed by reconciliation, opponents usually avoided each other for some time. These recomputed values are well below those of the three related or family wolf packs (44.1% to 53.2%) and in the range of the unrelated wolves at the Wolf Science Center.

4.2 Dogs at the Wolf Science Center

The second study was designed as a direct comparison (Cafazzo et al., 2018). Dogs from Hungarian shelters were hand-raised in peer groups from 10 days of age, like the wolves, and kept in four packs of two to four unrelated dogs (with one sibling pair). Over 403.26 hours of observation of the dogs between 2013 and 2015, each victim was filmed for 10 minutes after a conflict, and agreement between observers was high (κ = 0.82).

The dogs fought far less often than the wolves: 55 aggressive interactions, or 0.14 per hour, yielding 30 PC–MC pairs with 8 victims. Of these 30 conflicts, 28 were of high intensity (the paper's percentages for this split do not match the counts), and dog dyads had significantly more high-intensity conflicts than wolf dyads. Friendly contact between opponents occurred in 23.3% of PCs and 26.7% of MCs, and attracted and dispersed pairs did not differ (p = 0.59). There was no evidence of reconciliation. Instead, after conflicts, dog victims spent significantly less time near their aggressor than wolf victims, with no species difference in control periods. The authors concluded that "dog opponents do not seek an opportunity to reconcile but rather choose to avoid each other".

Three dogs had to be removed from their packs because conflicts escalated into severe aggression; no wolves were removed. The authors noted that this may underestimate how often dogs fight, but not their tendency to reconcile. The sample is small, and these were shelter-origin, hand-raised dogs in enclosures, not family dogs.

4.3 Why the Two Studies Disagree

Cafazzo and colleagues named two likely reasons. The first is relatedness and familiarity: the earlier groups included related dogs, and familiar dogs reconciled more, whereas the Wolf Science Center dogs were unrelated and assembled into packs. The wolf data point the same way, since the unrelated wolves had the lowest CCT of all wolf groups. The second is the ethogram described in Section 2.3. The authors concluded that reconciliation in dogs may depend on social and environmental conditions, such as relatedness, more than in wolves, and that its modulating factors "need to be further investigated". A review also describes the canine results as contrasting (Cordoni & Palagi, 2019). No third dog–dog study using this method could be found, and none exists for family dogs, daycare groups, shelters or free-ranging dogs. Everything said about reconciliation between family dogs is an extrapolation.

4.4 Avoidance as a Strategy

It is tempting to read the absence of reconciliation as a deficit. The comparative literature does not support that reading. Meerkats, a cooperatively breeding species, showed no reconciliation (a CCT of 0%) and used avoidance instead, as cited by Cafazzo et al. (2018) and Baan et al. (2014) from Kutsukake and Clutton-Brock (2008). Plausibly, avoidance reduces the risk of renewed aggression without a risky approach to the former opponent, although this has not been tested in dogs.

The pattern fits other comparisons from the same research center. In pair-wise food tests, subordinate dogs rarely challenged dominant partners, while in wolves both partners monopolized food and showed agonistic behavior to a similar extent; aggression was rare and weak in both species (Range et al., 2015). At a carcass, subordinate dogs mostly stayed away while the most dominant dog monopolized the food, whereas subordinate wolves spent as much time near and feeding from it as dominant ones; in dyadic tests, a strong affiliative bond promoted sharing in both species (Dale et al., 2017). As Cafazzo et al. (2018) discuss, dogs in these settings seem to keep their distance rather than negotiate (what comparisons of equally raised dogs and wolves show). Hand-raised wolves and dogs also behaved similarly toward a territorial threat and during separation from the pack, but only wolves showed related hormonal changes, which the authors interpreted as stronger emotional activation to threats to group integrity (Wirobski et al., 2023). That is not a conflict study, but it is consistent with group cohesion weighing less for dogs in such packs.

Taken together, avoidance and distance-keeping after a conflict may be the normal strategy of dogs rather than a failure. The data are few and do not come from family dogs, but nothing in the literature suggests that a dog that keeps away from its opponent after a squabble is behaving abnormally.

4.5 Ritualized Conflict in Free-Ranging and Daycare Groups

Much conflict management in dogs may happen before or instead of fighting. In a group of free-ranging dogs, the dominance hierarchy was linear, rank correlated with age, submissive–affiliative behavior met the criteria for a formal submission signal, although not all dogs showed it, and high-ranking dogs generally had priority access to food (Cafazzo et al., 2010). In five packs of free-ranging and semi-free-ranging dogs, "potentially injurious aggression was very rare", submissive reversals were more common than in wolves, and the authors suggested that rank is usually not acquired through fighting (Bonanni et al., 2017). In a daycare group of 24 neutered companion dogs, submission was a better status indicator than aggression, aggression consisted mainly of ritualized threats, and conflicts with contact were never injurious; dogs could disengage and move away (Trisko & Smuts, 2015). In a kennel group of 16 intact dogs, lowering of posture was the best overall status indicator (van der Borg et al., 2015).

These studies use dominance descriptively, for the asymmetry of agonistic relationships within a pair. They do not support the idea that dogs fight to gain rank or that a fight settles a rank question after which peace returns; no post-conflict study in dogs tested that. Why dominance fails as an explanation of dog behavior is discussed elsewhere (what the dominance concept can and cannot explain), as is the limited transfer from free-ranging to pet dogs (how free-ranging dogs organize their social life).

5. Third Parties: Affiliation, Consolation and Interpretation

5.1 Bystanders in Wolves

In the Pistoia pack, with 580 PC–MC pairs, researchers distinguished unsolicited third-party affiliation, from a bystander to the victim, from solicited affiliation, from the victim to a bystander (Palagi & Cordoni, 2009). Both exceeded baseline in the first two or three minutes after a conflict. Unsolicited contacts exceeded baseline after low- and medium-intensity conflicts but not after high-intensity aggression, which the authors read as bystanders avoiding potentially harmful situations. They were more frequent between close partners and when no reconciliation had occurred, and renewed aggression was lower when they occurred. Victims that received more unsolicited contact performed less redirected aggression. The authors cautioned that the cognitive skills behind conflict resolution in canids "still have to be investigated in detail". A later analysis found that affiliation toward the victim reduced renewed attacks by the former aggressor, while affiliation toward the aggressor limited attacks on uninvolved group members (Cordoni & Palagi, 2015).

5.2 Bystanders in Dogs

For dogs, the only data come from Cools et al. (2008). As summarized in a review, bystander contacts were more often directed to victims than to aggressors, victim-directed affiliation was "markedly higher" than affiliation between former opponents, and familiarity did not seem to affect it significantly; the data were too scarce to investigate the functions of dog "consolation" and "appeasement" (Cordoni & Palagi, 2019). Ravens show a similar pattern, with weak evidence for reconciliation and strong evidence for bystander affiliation (Fraser & Bugnyar, 2010). Whether dogs reliably show it has not been tested beyond the one descriptive study.

5.3 Why Consolation Is an Interpretive Label

Bystander affiliation toward a victim is often called consolation, a label that implies the contact reduces the victim's distress. That has to be shown. Even in chimpanzees, the stress-alleviating effect had been assumed but "never been demonstrated" until a study showed reduced behavioral stress measures in recipients (Fraser et al., 2008). In chimpanzees at Chester Zoo, consolation and reconciliation tended to substitute for each other (Fraser & Aureli, 2008). Whether consolation reflects empathy is contested: in a large chimpanzee dataset, it occurred mainly between socially close individuals, and high-ranking males often consoled, probably as part of policing (Romero et al., 2010). Japanese macaques showed reconciliation but no consolation, even among kin (Aureli et al., 1993), so its presence has to be tested species by species.

Outside primates, prairie voles groomed familiar cagemates more after those had experienced a stressor, matched their fear and corticosterone increase, and stopped doing so when oxytocin receptors in the anterior cingulate cortex were blocked (Burkett et al., 2016). This is neither a post-aggression study nor about dogs, but it shows what evidence would be needed. For dog–dog conflicts, none exists. When a third dog walks over to a dog that has just been snapped at, the observation is that a bystander approached the victim. Whether it is comforting, checking, seeking contact for itself or passing through is not known.

5.4 Responses to a Conspecific's Distress

The closest dog study does not involve conflict. In 16 pairs of cohabiting dogs, playback of whines increased alertness and stress-related behaviors compared with a control sound, and dogs then showed more comfort-offering behavior toward their familiar partner (Quervel-Chaumette et al., 2016). The authors noted that dogs might have responded to a dog versus a non-dog sound rather than to emotional content. The study shows a response to a conspecific's distress signal, not consolation after a fight (what emotional contagion studies do and do not show). The calming effect of a familiar partner is usually studied as social buffering, a different concept (how a familiar partner can buffer stress).

6. Conflicts with Humans: A Separate Question

6.1 After Being Scolded

In one study, an owner scolded the dog for "stealing" human food while a second owner stayed passive and an experimenter filmed (Cavalli et al., 2016). The dog's behavior was recorded for 3 minutes 30 seconds before and after. Of 24 dogs tested, 17 adult dogs aged 1 to 12 years remained; dogs with aggression or excessive fear toward humans had been excluded. After the scolding, dogs increased closeness, gazing and tail wagging toward the scolding owner and showed appeasement behaviors such as averting the eyes, lowered ears, lip licking and crouching. They did not seek more contact with the second owner, so there was no evidence of consolation-seeking. The authors noted that a stranger was present, that most of these behaviors occurred within 30 seconds, that only the dog could initiate contact and that only a foraging context was tested. The behaviors overlap with those in research on the so-called guilty look, which the authors discuss (what the guilty look studies found).

6.2 When Owners Argue

In two further studies, dogs watched their owners argue (Rial et al., 2025). In both, dogs directed more affiliation at the "victim" of the argument than in the control condition. More stress-related behaviors appeared only in the second study, in which dogs also directed behaviors at the "aggressor" "that could be interpreted as appeasement". The bond to each owner did not appear to modulate affiliation, and effects on the person's emotional state remain to be studied. Sample sizes are not available from the sources used here.

6.3 Why This Does Not Answer the Dog–Dog Question

These studies show that dogs respond to conflict involving their people with approach, affiliation and appeasement-like behavior. A dog that approaches a scolding owner may be responding to the owner's change in behavior, to what it expects next or to the relationship. None of this can be transferred to fights between dogs, for which the only data come from captive groups.

7. Relationship Quality and Familiarity

7.1 The Valuable Relationship Hypothesis

The most influential explanation is the valuable relationship hypothesis: animals should reconcile more with partners whose relationship is worth preserving. De Waal (2000) argued that the full set of checks on aggression can be expected "only when social relationships are valued". In the classic experiment, researchers raised the value of long-tailed macaque pair relationships by training a cooperative feeding task, and reconciliation increased in 6 of 7 dyads, with a median rate three times the baseline (Cords & Thurnheer, 1993). The authors noted that "there are other potential explanations". Seven primate dyads are not evidence about dogs.

7.2 Competing Explanations

The integrated hypothesis holds that conflicts between close partners cause more post-conflict anxiety, which makes reconciliation more likely. It was mostly supported in chimpanzees at Arnhem (Koski et al., 2007) but not in the Arctic wolves (Lazzaroni et al., 2017). Another view is that reconciliation serves short-term goals such as access to resources. A primate review warned: "There are good reasons to question the validity of the relationship-repair model, but it remains firmly entrenched in the reconciliation literature, perhaps because it fits our own folk model" (Silk, 2002). The warning applies with particular force to dogs.

7.3 What the Canid Data Show

In wolves, relationship quality predicted reconciliation weakly or not at all. In Pistoia, there was only a trend (CCT 56.0% in weak vs 74.1% in close relationships, p < 0.1) and no group-level correlation (Cordoni & Palagi, 2008). At the Wolf Science Center, symmetry of affiliation had no effect (Cafazzo et al., 2018). The Arctic wolf authors suggested that high relatedness may reduce the importance of relationship quality (Lazzaroni et al., 2017). In dogs, the only relevant finding is that familiar dogs reconciled more in Cools et al. (2008), with familiarity and relatedness largely confounded.

A review went further: "The social repairing function ascribed to the reconciliation mechanism is probably useless for dogs", linked to the absence of cooperative hunting and collective rearing (Cordoni & Palagi, 2019). This is the authors' interpretation, not a tested finding. It fits the avoidance data, but companion dogs living closely with one another and with people have not been studied.

8. Preventing Conflict Before It Starts

8.1 Play: Asymmetry Without Collapse

A popular rule holds that dogs keep play friendly by taking turns as winner and loser. In 55 dyads in a dog park, only 5.4% were balanced in these roles, and 21.8% were completely asymmetric (Bauer & Smuts, 2007). In 203 play sessions, asymmetry shortened sessions, which the authors interpreted as "limiting the risk of possible escalation" (Cordoni et al., 2016). Play bows mostly followed pauses and re-initiated play rather than marking risky actions (Byosiere et al., 2016), and rapid mimicry of play signals went with longer sessions and stronger bonds (Palagi et al., 2015). Dogs seem to regulate play through pauses, re-initiation and ending early rather than through strict fairness (how play differs from escalation).

8.2 Calming and Appeasement Signals, Read in Context

Calming signals are widely taught as behaviors that stop conflict. In the only dog–dog pilot study, 24 dogs each met four dogs off leash (Mariti et al., 2017). None of 109 aggressive episodes was preceded by a calming signal. After receiving aggression, senders showed at least one calming signal in 67.0% (73) of cases; in 79.4% of these (58), the aggression de-escalated, in 5.5% it increased and in 15.1% it stayed unchanged. Muzzle licking was more frequent with familiar dogs. There was no comparison of how often aggression de-escalates without such signals, and the categories were not individually validated. When unfamiliar people approached 53 dogs in a mildly threatening or neutral way, nose licking and lip wiping went with a non-reactive attitude regardless of condition; the authors concluded that putative appeasement signals were "not strictly associated with a conflict-ridden situation" (Pedretti et al., 2023) (what studies on calming signals support).

8.3 Submission, Distance and Context

Formal submission signals and moving away are further forms of prevention. In the daycare group, muzzle licks met criteria for a formal submission display except for coverage (Trisko & Smuts, 2015), and in the kennel group, body tail wagging and low postures were the best submission indicators (van der Borg et al., 2015). None of these behaviors has a fixed meaning. What a lick, a head turn, a sniff or a step away means depends on who is involved (age, sex, familiarity, history), what the conflict was about (food, a resting place, a person, play that tipped over), how the two dogs usually relate and the setting (room to move away, other dogs, people). Signals and post-conflict behavior can only be read in this context, and even then the reading is a hypothesis to be checked against what happens next. "He is apologizing" goes beyond anything that has been measured.

9. Multi-Dog Households and Second Dogs

9.1 What Clinical and Survey Data Show

Data on conflicts between dogs at home come from referral cases and surveys. In 38 referral pairs, 79% were same-sex, and where an instigator was identified, it was mostly the younger (20 of 27) or newer dog (19 of 27). Triggers included owner attention, food, excitement and found items. In 15 of 30 cases with injury data, a dog needed veterinary treatment, and in 3 an owner needed medical attention. Owners reported 69% overall improvement after treatment, and the authors concluded that "consistency and predictability of social interactions are essential" (Wrubel et al., 2011). In 305 referral pairs, resource guarding was the most common trigger (72.8%), aggressors were on average 16 months younger, and a poor outcome was associated with same-sex pairs, bites that broke the skin and aggression on sight (Feltes et al., 2020) (what is known about resource guarding). In a UK survey, 22% of owners reported current aggression toward unfamiliar dogs and 8% toward household dogs; household aggression was associated with increasing age and with the use of positive punishment or negative reinforcement, among other factors; the models explained less than 15% of the variance (Casey et al., 2013). These are associations from problem cases, not risk estimates for every pair (what research shows about conflict in multi-dog homes). No study supports leaving fights to run their course.

9.2 Give Space Instead of Forcing Contact

If avoidance after conflict is a normal strategy for dogs, distance after a squabble is not something that needs fixing, and no study shows that bringing dogs together afterwards helps.

In practice: After a squabble, we advise owners to give both dogs room rather than bringing them together to "make up": separate resting places, no shared treats or close handling of both dogs at once for a while, and letting the dogs decide when they approach each other again. This comes from our training practice; it has not been tested in a study.

9.3 Management in Multi-Dog Homes

In practice: In homes with more than one dog, we set things up so that dogs do not have to compete: separate feeding places, high-value chews only when the dogs are apart, several resting places with distance between them, and gates or closed doors in busy moments such as arrivals or meal preparation. This is practical experience that addresses the triggers in the case series, not a tested protocol.

9.4 Watch for Repeated Conflicts Between the Same Pair

In practice: We ask owners to note when, where and between which dogs conflicts happen, what came before and how long the dogs kept apart afterwards. If the same pair clashes repeatedly, if conflicts become more frequent or intense, or if one dog starts avoiding certain rooms, we treat that as a reason for a veterinary check and a professional assessment rather than waiting for the dogs to settle it. This is our working approach, not a validated method (how aggression is understood by its function).

9.5 Interrupting a Fight

Shouting, physical punishment or other aversive interruption can stop a fight in the moment. The question is what the dogs learn from it and at what cost: whether the other dog comes to predict something unpleasant, whether fear rises in an already tense situation, and the fact that the trigger is still there afterwards (what is known about the side effects of punishment). The household series also show that owners can be injured (Wrubel et al., 2011).

In practice: We advise separating fighting dogs with something between them, such as a door, a gate or a board, rather than with hands, and keeping them apart until they have calmed down and the situation has been assessed. For a second dog, we look at how resources, space and attention will be shared, and how easily each dog can keep its distance. This comes from practice, not from a study.

10. Summary at a Glance

Post-conflict behavior is measured by comparing a period after a conflict with a matched control. Reconciliation is shown only when friendly contact between former opponents is earlier or more frequent after conflicts (Veenema et al., 1994), and results depend on behavior definitions, time windows and context. Wolves reconciled in all groups studied, with CCTs from 22.46% to 53.2%, mostly within one or two minutes and less often over food (Cordoni & Palagi, 2008; Baan et al., 2014; Lazzaroni et al., 2017; Cafazzo et al., 2018). Functional evidence is limited to less renewed and redirected aggression in one captive pack; no canid study has shown stress reduction.

In dogs, the evidence is split and thin: one study of three captive groups reported reconciliation (Cools et al., 2008), while the only matched wolf–dog comparison found avoidance (Cafazzo et al., 2018). Avoidance may be the normal strategy. "Consolation" between dogs is an interpretive label without a functional test. Studies of conflicts with people say nothing direct about dog–dog conflict (Cavalli et al., 2016; Rial et al., 2025). Post-conflict behavior can only be read in context.

11. Research Gaps and Critical Appraisal

11.1 Few Studies, Small Samples

For dog–dog conflicts, there are two PC–MC studies with opposite results. The one that found no reconciliation had 30 PC–MC pairs and 8 victims (Cafazzo et al., 2018). The wild wolf data comprise 34 conflicts from two packs in one winter (Baan et al., 2014), and each zoo study concerns a single pack (Cordoni & Palagi, 2008; Lazzaroni et al., 2017). Replication is almost entirely missing.

11.2 Populations

No dog–dog study involved family dogs. One used captive groups of unclear origin and housing, the other shelter-origin dogs hand-raised with unrelated pack members in enclosures. Kinship, familiarity and living conditions differ fundamentally from a typical two-dog household.

11.3 Definitions and Measures

Ethograms, time windows and control rules differ between studies, and broad definitions may inflate both conflicts and reconciliations (Cafazzo et al., 2018). In one key paper, reported percentages do not match the counts for two results. Stress indicators after conflict are poorly validated for canids: in Arctic wolves, lip licking dominated, while scratching, the classic primate measure, was almost absent (Lazzaroni et al., 2017).

11.4 Function and Secondary Data

The only significant functional result in canids comes from one captive pack (Lazzaroni et al., 2017), and no study tested the function of third-party affiliation in dogs. The valuable relationship hypothesis rests experimentally on seven macaque dyads (Cords & Thurnheer, 1993) and is questioned within primatology (Silk, 2002). The human-directed studies had small or unreported samples and a single context (Cavalli et al., 2016; Rial et al., 2025). The numbers for the first dog study come from later papers, and several primate sources are cited only for their concepts.

12. Conclusion

Reconciliation is well documented in primates and wolves, and the method for measuring it is well established. For dogs, one study of captive groups reported reconciliation, and the only direct comparison with equally raised wolves found that dogs avoided each other after conflicts. Avoidance and distance-keeping may be the normal strategy of dogs rather than a sign that something has gone wrong. Claims that dogs make up, apologize or console each other after a fight have not been tested in a way that supports them, and a stress-reducing function of reconciliation has been shown only in primates.

For owners, this shifts the focus from waiting for, or arranging, a reconciliation to preventing conflicts through management, giving dogs room after a squabble and watching for patterns between the same dogs. What a lick, a head turn or a step away means can only be judged in context: who was involved, what the conflict was about, how the dogs usually relate and where it happened.

Key Insights (Takeaways)

• Reconciliation is shown only when friendly contact between former opponents is earlier or more frequent after a conflict than in a matched control (Veenema et al., 1994).

• Wolves reconciled in all packs studied, mostly within one or two minutes and less often after conflicts over food (Cordoni & Palagi, 2008; Lazzaroni et al., 2017; Cafazzo et al., 2018).

• In dogs, the evidence is split: one study of captive groups reported reconciliation (Cools et al., 2008), while the only matched wolf–dog comparison, with 30 PC–MC pairs, found avoidance (Cafazzo et al., 2018).

• Keeping distance after a conflict may be a normal dog strategy, not a failure of social skills.

• Calling a third dog's approach "consolation" is an interpretation; stress reduction through consolation or reconciliation has been shown only in primates (Fraser et al., 2008).

• Studies of conflicts with people show approach and appeasement-like behavior toward humans but say nothing direct about fights between dogs (Cavalli et al., 2016; Rial et al., 2025).

• Signals and post-conflict behavior can only be read in context: who, what conflict, which relationship, which setting.

• In multi-dog homes, managing triggers, giving space after a squabble and watching for repeated conflicts between the same pair are more useful than expecting the dogs to make up.

References

Aureli, F., Cords, M., & van Schaik, C. P. (2002). Conflict resolution following aggression in gregarious animals: A predictive framework. Animal Behaviour, 64(3), 325–343. https://doi.org/10.1006/anbe.2002.3071

Aureli, F., & van Schaik, C. P. (1991). Post-conflict behaviour in long-tailed macaques (Macaca fascicularis): II. Coping with the uncertainty. Ethology, 89(2), 101–114. https://doi.org/10.1111/j.1439-0310.1991.tb00297.x

Aureli, F., Veenema, H. C., van Panthaleon van Eck, C. J., & van Hooff, J. A. R. A. M. (1993). Reconciliation, consolation, and redirection in Japanese macaques (Macaca fuscata). Behaviour, 124(1–2), 1–21. https://doi.org/10.1163/156853993X00470

Baan, C., Bergmüller, R., Smith, D. W., & Molnar, B. (2014). Conflict management in free-ranging wolves, Canis lupus. Animal Behaviour, 90, 327–334. https://doi.org/10.1016/j.anbehav.2014.01.033

Bauer, E. B., & Smuts, B. B. (2007). Cooperation and competition during dyadic play in domestic dogs, Canis familiaris. Animal Behaviour, 73(3), 489–499. https://doi.org/10.1016/j.anbehav.2006.09.006

Bonanni, R., Cafazzo, S., Abis, A., Barillari, E., Valsecchi, P., & Natoli, E. (2017). Age-graded dominance hierarchies and social tolerance in packs of free-ranging dogs. Behavioral Ecology, 28(4), 1004–1020. https://doi.org/10.1093/beheco/arx059

Burkett, J. P., Andari, E., Johnson, Z. V., Curry, D. C., de Waal, F. B. M., & Young, L. J. (2016). Oxytocin-dependent consolation behavior in rodents. Science, 351(6271), 375–378. https://doi.org/10.1126/science.aac4785

Byosiere, S.-E., Espinosa, J., & Smuts, B. (2016). Investigating the function of play bows in adult pet dogs (Canis lupus familiaris). Behavioural Processes, 125, 106–113. https://doi.org/10.1016/j.beproc.2016.02.007

Cafazzo, S., Marshall-Pescini, S., Lazzaroni, M., Virányi, Z., & Range, F. (2018). The effect of domestication on post-conflict management: Wolves reconcile while dogs avoid each other. Royal Society Open Science, 5(7), 171553. https://doi.org/10.1098/rsos.171553

Cafazzo, S., Valsecchi, P., Bonanni, R., & Natoli, E. (2010). Dominance in relation to age, sex, and competitive contexts in a group of free-ranging domestic dogs. Behavioral Ecology, 21(3), 443–455. https://doi.org/10.1093/beheco/arq001

Casey, R. A., Loftus, B., Bolster, C., Richards, G. J., & Blackwell, E. J. (2013). Inter-dog aggression in a UK owner survey: Prevalence, co-occurrence in different contexts and risk factors. Veterinary Record, 172(5), 127. https://doi.org/10.1136/vr.100997

Cavalli, C., Dzik, V., Carballo, F., & Bentosela, M. (2016). Post-conflict affiliative behaviors towards humans in domestic dogs (Canis familiaris). International Journal of Comparative Psychology, 29. https://doi.org/10.46867/ijcp.2016.29.00.03

Cools, A. K. A., Van Hout, A. J.-M., & Nelissen, M. H. J. (2008). Canine reconciliation and third-party-initiated postconflict affiliation: Do peacemaking social mechanisms in dogs rival those of higher primates? Ethology, 114(1), 53–63. https://doi.org/10.1111/j.1439-0310.2007.01443.x

Cordoni, G., Nicotra, V., & Palagi, E. (2016). Unveiling the "secret" of play in dogs (Canis lupus familiaris): Asymmetry and signals. Journal of Comparative Psychology, 130(3), 278–287. https://doi.org/10.1037/com0000035

Cordoni, G., & Palagi, E. (2008). Reconciliation in wolves (Canis lupus): New evidence for a comparative perspective. Ethology, 114(3), 298–308. https://doi.org/10.1111/j.1439-0310.2008.01474.x

Cordoni, G., & Palagi, E. (2015). Being a victim or an aggressor: Different functions of triadic post-conflict interactions in wolves (Canis lupus lupus). Aggressive Behavior, 41(6), 526–536. https://doi.org/10.1002/ab.21590

Cordoni, G., & Palagi, E. (2019). Back to the future: A glance over wolf social behavior to understand dog–human relationship. Animals, 9(11), 991. https://doi.org/10.3390/ani9110991

Cords, M., & Thurnheer, S. (1993). Reconciling with valuable partners by long-tailed macaques. Ethology, 93(4), 315–325. https://doi.org/10.1111/j.1439-0310.1993.tb01212.x

Dale, R., Range, F., Stott, L., Kotrschal, K., & Marshall-Pescini, S. (2017). The influence of social relationship on food tolerance in wolves and dogs. Behavioral Ecology and Sociobiology, 71(7), 107. https://doi.org/10.1007/s00265-017-2339-8

de Waal, F. B. M. (2000). Primates – A natural heritage of conflict resolution. Science, 289(5479), 586–590. https://doi.org/10.1126/science.289.5479.586

de Waal, F. B. M., & van Roosmalen, A. (1979). Reconciliation and consolation among chimpanzees. Behavioral Ecology and Sociobiology, 5(1), 55–66. https://doi.org/10.1007/BF00302695

de Waal, F. B. M., & Yoshihara, D. (1983). Reconciliation and redirected affection in rhesus monkeys. Behaviour, 85(3–4), 224–241. https://doi.org/10.1163/156853983X00237

Feltes, E. S. M., Stull, J. W., Herron, M. E., & Haug, L. I. (2020). Characteristics of intrahousehold interdog aggression and dog and pair factors associated with a poor outcome. Journal of the American Veterinary Medical Association, 256(3), 349–361. https://doi.org/10.2460/javma.256.3.349

Fraser, O. N., & Aureli, F. (2008). Reconciliation, consolation and postconflict behavioral specificity in chimpanzees. American Journal of Primatology, 70(12), 1114–1123. https://doi.org/10.1002/ajp.20608

Fraser, O. N., & Bugnyar, T. (2010). Do ravens show consolation? Responses to distressed others. PLoS ONE, 5(5), e10605. https://doi.org/10.1371/journal.pone.0010605

Fraser, O. N., Stahl, D., & Aureli, F. (2008). Stress reduction through consolation in chimpanzees. Proceedings of the National Academy of Sciences, 105(25), 8557–8562. https://doi.org/10.1073/pnas.0804141105

Koski, S. E., Koops, K., & Sterck, E. H. M. (2007). Reconciliation, relationship quality, and postconflict anxiety: Testing the integrated hypothesis in captive chimpanzees. American Journal of Primatology, 69(2), 158–172. https://doi.org/10.1002/ajp.20338

Kutsukake, N., & Clutton-Brock, T. H. (2008). Do meerkats engage in conflict management following aggression? Reconciliation, submission and avoidance. Animal Behaviour, 75(4), 1441–1453. https://doi.org/10.1016/j.anbehav.2007.09.018

Lazzaroni, M., Marshall-Pescini, S., & Cafazzo, S. (2017). Post-conflict opponent affiliation reduces victim re-aggression in a family group of captive arctic wolves (Canis lupus arctos). PLOS ONE, 12(11), e0187450. https://doi.org/10.1371/journal.pone.0187450

Mariti, C., Falaschi, C., Zilocchi, M., Fatjó, J., Sighieri, C., Ogi, A., & Gazzano, A. (2017). Analysis of the intraspecific visual communication in the domestic dog (Canis familiaris): A pilot study on the case of calming signals. Journal of Veterinary Behavior, 18, 49–55. https://doi.org/10.1016/j.jveb.2016.12.009

Palagi, E., & Cordoni, G. (2009). Postconflict third-party affiliation in Canis lupus: Do wolves share similarities with the great apes? Animal Behaviour, 78(4), 979–986. https://doi.org/10.1016/j.anbehav.2009.07.017

Palagi, E., Nicotra, V., & Cordoni, G. (2015). Rapid mimicry and emotional contagion in domestic dogs. Royal Society Open Science, 2(12), 150505. https://doi.org/10.1098/rsos.150505

Pedretti, G., Canori, C., Biffi, E., Marshall-Pescini, S., & Valsecchi, P. (2023). Appeasement function of displacement behaviours? Dogs' behavioural displays exhibited towards threatening and neutral humans. Animal Cognition, 26(3), 943–952. https://doi.org/10.1007/s10071-023-01742-9

Quervel-Chaumette, M., Faerber, V., Faragó, T., Marshall-Pescini, S., & Range, F. (2016). Investigating empathy-like responding to conspecifics' distress in pet dogs. PLOS ONE, 11(4), e0152920. https://doi.org/10.1371/journal.pone.0152920

Range, F., Ritter, C., & Virányi, Z. (2015). Testing the myth: Tolerant dogs and aggressive wolves. Proceedings of the Royal Society B, 282(1807), 20150220. https://doi.org/10.1098/rspb.2015.0220

Rial, L. A., Cavalli, C., Dzik, M. V., & Bentosela, M. (2025). Third-party affiliation in domestic dogs during and after a human conflict. Ethology, 131(1), e13522. https://doi.org/10.1111/eth.13522

Romero, T., Castellanos, M. A., & de Waal, F. B. M. (2010). Consolation as possible expression of sympathetic concern among chimpanzees. Proceedings of the National Academy of Sciences, 107(27), 12110–12115. https://doi.org/10.1073/pnas.1006991107

Silk, J. B. (2002). The form and function of reconciliation in primates. Annual Review of Anthropology, 31(1), 21–44. https://doi.org/10.1146/annurev.anthro.31.032902.101743

Trisko, R. K., & Smuts, B. B. (2015). Dominance relationships in a group of domestic dogs (Canis lupus familiaris). Behaviour, 152(5), 677–704. https://doi.org/10.1163/1568539X-00003249

van der Borg, J. A. M., Schilder, M. B. H., Vinke, C. M., & de Vries, H. (2015). Dominance in domestic dogs: A quantitative analysis of its behavioural measures. PLOS ONE, 10(8), e0133978. https://doi.org/10.1371/journal.pone.0133978

Veenema, H. C., Das, M., & Aureli, F. (1994). Methodological improvements for the study of reconciliation. Behavioural Processes, 31(1), 29–37. https://doi.org/10.1016/0376-6357(94)90035-3

Wirobski, G., Range, F., Graat, E. A. M., Palme, R., Deschner, T., & Marshall-Pescini, S. (2023). Similar behavioral but different endocrine responses to conspecific interactions in hand-raised wolves and dogs. iScience, 26(2), 105978. https://doi.org/10.1016/j.isci.2023.105978

Wrubel, K. M., Moon-Fanelli, A. A., Maranda, L. S., & Dodman, N. H. (2011). Interdog household aggression: 38 cases (2006–2007). Journal of the American Veterinary Medical Association, 238(6), 731–740. https://doi.org/10.2460/javma.238.6.731