Reconciliation in Primates and Wolves
Ali and Jane are kicking around some ideas about how the concept of reconciliation, which was first proposed for chimpanzees (DeWaal & VanRoosmalen 1979) has been applied to wolves….
Ali and Jane are kicking around some ideas about how the concept of reconciliation, which was first proposed for chimpanzees (DeWaal & VanRoosmalen 1979) has been applied to wolves….
August 13th, 2009 at 10:25 am
Ali is telling me that when she went back and reread the article by DeWaal, it made more sense than when she read about the concept in the recent article on wolves by Cordoni & Palagi (2008). In the original, it was a specific set of behaviors, i.e. kissing and hugging, that occurred in a short time frame after the conflict and broke the tension between the rivals.
In the later studies of both wolves and chimps, the behaviors are not described as specifically. The current definition of reconciliation is operational, simply a statistical thing. “If we see more attracted than dispersed pairs, then we agree to call it reconciliation”.
Lets pretend that wolves reconcile by muzzle lick. That was not recorded, we are not capturing it in a sample taken 24 hours later. This is the protocol for the matched control (MC). The post conflict (PC) period is from t =0 for 30 or 45 min. They would have recorded any contact during the PC period, and did not report the actual behaviors that they observed. They would have recorded which behaviors occurred, but did not report it. Its hard for us to decide if we agree with their interpretation without that evidence.
August 13th, 2009 at 10:40 am
Seems to me that there is a problem with making explicit what are the cause and effect variables. In my mind, the reconciliation interaction would be a cause variable. The nature of the relationship (disperse or attracted) would be an effect variable.
The way they measured the relationship was by comparing the data in the PC and MC periods. If they had affinitive contact sooner in the PC than the MC then it was interpretted as attractive, otherwise dispersed.
For example, an instance of attracted pair would be in the following scenario:
If A& B fight, then five minutes later body rub…then the following day in the matched control period A&B don’t body rub until 10 minutes into the observation period, then this instance would be evidence of an attracted pair.
The equation for CCT is something like attracted minus dispersive over the total number of interactions. Translating, this is an interpretation of the probability of attracted instances.
This was from the protocol published by Veenema et al. 1984 for a metric that they used to compare the corrected conciliatory tendency (CCT) for individuals. So they pool the info over all instances and relationships for one individual
August 13th, 2009 at 10:43 am
In Ali’s draft paper, we need to be much more concise. We could condense the primate info and expand the canid info.
We could condense some of the information about the dog study.
Aureli et al. (200?) introduces the theoretical framework for the valuable relationship hypothesis. What this does is providing a basis for predicting when reconciliation would be more likely. It was handy, because it gave a more general hypothesis that could be tested across species and was not specific to particular behaviors.
So we have a tension between testing generalities (good science) and needing to know the specifics of how to interpret behaviors in each species (essential to doing good science).
August 13th, 2009 at 11:32 am
I have a problem with how we define “valuable relationship”. As I understood, a relationship with a dominant is more valuable than a relationship with a subordinate.
However, from an ultimate perspective, valuable should be in terms of “what helps me make more copies of my genotype”. In wolves, dominance interactions may depend more on which individual first got access to the resource. The relationships that are important are between care givers and care receivers. That would map directly onto ultimate fitness, because if I get care, I am more likely to survive to reproduce. If I give care to my own offspring or close relatives, my genotype is more likely to make more copies of itself.
Correct me if I am wrong, but the measure of the value of a relationship was based on the number of aggressive actions emitted vs. those received, was it not?
We actually have data on these relationships in my doctoral dissertation. However, those chapters were never published. Maybe I need to dust them off?
August 13th, 2009 at 11:56 am
Ali, here is a copy of the manuscript you wrote for our Vertebrate Behavior course in Spring 2009. This is what gave us the idea to kick around the ideas and write a thoughtful commentary, hopefully for publication in Ethology.
http://wfsc.tamu.edu/jpackard/scienceinaction/pred_Krzton090813.pdf