Bill, as you requested, here's some feedback on your article. Please don't mistake the clinical tone of it as anything other than constructive. I can see what you are exploring, and it's really straightforward if one is familiar with the concepts and terminology of learning and behavior.
Just to forewarn the rest of you, if you don't like my previous posts and my approach, save yourself the trouble and don't bother reading this, as it's more "over-complication," as some of you would put it.
*My comments are underlined and italicized*
Despite a rather romantic notion of wild hawks being capable of developing some sort of emotional attachment to falconers, there is little doubt that passage hawks return to us after only a few weeks of training, and continue to return to us throughout the season because of conditioning, with food being the reinforcement reinforcing stimulus. Just nit picking. And, because of this conditioning, we know that these hawks can be released back into the wild and yet return to the fist days later; for example, Tasha Leong released a passage red-tail only to have it return to her fist, having not seen it for eleven days. Eric Fontaine had a passage goshawk remain in area and return to his home multiple times over a two year period. He called her to the lure and fed her off the fist regularly.
It seems the longer they are in captivity, the more profound the conditioning. Like Eric, another falconer released a four time intermewed RT on his property at the end of the hawking season. The hawk remained in the area, and he was able to pick it up off a fence post in the fall, and hawk it for another two seasons. Charles Warwick released a three-time intermewed red-tail in his back yard, and it also hung around for months; in fact, it continued to hunt with Charles when he took his dogs for a walk, and eventually took on a mate in the area.
But, even more intriguing are the numerous incidents when passage hawks have been released miles from their mews and yet find their way home. For example, Dana Brenfleck had a passage red-tail return home after being released 30 miles away. Charles lost the previously mentioned red-tail 25 miles from home. Three months later it showed up at his house, and came right to the fist. There are enough similar stories to ask why these passage hawks continue to return to the fist and seek familiar distant territory? Can conditioning be that strong and lasting, or is there something beyond conditioning that prompts them to return? This article will explore these questions. Behavior that is repeated is being reinforced, or conditioned. There isn’t anything “beyond” conditioning, so I’m not sure what the premise of the article is at this point...
These associative behaviors suggest the presence of a certain amount of intelligence; and, for animal intelligence to be present, there must be cognition. According to Sara Shettlesworth, “cognition refers to the mechanisms by which animals acquire, process, store, and act on information from the environment. These include perception, learning, memory, and decision making" (Cognition, Evolution, and Behavior (5)). Just a general comment: all animals have cognition and learn, including simple organisms like insects and even slime mold. Since it’s clear, through their behavior, that passage hawks perceive, learn, remember and make decisions, they demonstrate a certain level of cognition, and, therefore, a degree of intelligence that negates the idea that these birds return to us because “they are dumber than a rock.” They make a clear decision to return to a distant location or to the fist, particularly after being free for any length of time. But, again, the question is why, particularly after being free long enough to realize that they are on their own, and fully capable of surviving?
Much of what a hawk learns in the wild comes from conditioning brought on by “associative learning.” Associative learning is based on the assumption that life experiences reinforce one another and can be linked to enhance the learning process. So, a red-tailed hawk begins to associate rats brought to it by its parents with food. This association carries over once it fledges, and it begins to catch rats in order to survive. The process of pursuit and catching becomes linked (through association) with what the rat represents—food. And for lower level intelligence, such as with raptors, associative learning is paramount for survival, particularly since it must catch its food in order to survive. I personally wouldn’t regard the importance of associative learning as more or less important when it comes to general “intelligence.” Associative learning is the ability of an organism to discriminate and generalize like and unlike stimuli, and that correlates directly to intelligence, so I’m not really following this thought. If associative learning were indeed more important to a species like a raptor that has to catch its own food, than say, a parrot or corvid that forages, then a raptor would have developed superior associative learning skills. As a result, hawks, as predators, are very receptive to associative learning (particularly if it has to do with food), which is central to the conditioning process used by falconers during initial training, and is the main reason why passage hawks respond so quickly. Again, I don’t follow. You’re saying that hawks in general are innately receptive to associative learning, but for some reason passage hawks are more so than a captive-bred hawk? A good example of how receptive hawks are to associative learning and how much it impacts their behavior is demonstrated in another experience by Eric Fontaine. A couple of passage goshawks were decimating his pigeons, so he trapped them and released them two mountain ranges and over 100 miles away from his home. He flagged them prior to release. One returned to his home seven days later and the other ten. So, again, this associative learning, particularly when it involves food, is very significant to a hawk, whether it comes about by virtue of a falconer during the training process, or not. Is it really remarkable that a hawk will return to a proven food resource? Many animals of far lower intelligence will do the same.
In terms of training a hawk for the purposes of falconry, two types of conditioning need to be identified; the first of which is often attributed to training hawks—“operant conditioning.” Operant condition refers to the use of consequences to modify behavior. The classic example is rewarding a rat with food for pushing a lever. The food, of course, is the “reinforcer.” So, when a hawk comes to the fist and is rewarded with food for doing so, operant conditioning is being used.
The other form of conditioning is called “classical,” or “Pavlovian conditioning.” This form of conditioning attempts to get a desired response by using a stimuli that is unrelated to the target behavior; for example, Pavlov got dogs to salivate simply by ringing a bell. To a certain extent, an example of classical conditioning in falconry would be conditioning a hawk to come to the fist based upon a whistle, rather than food, but the predominant type of conditioning used by falconers is operant in nature through the use of food as reward. This isn’t correct. Classical conditioning is reflexive behavior that occurs without any decision making. Yes, the whistle can elicit salivation in hawks, just as with dogs, and be associated with food, but if you whistle and the hawk flies to the glove even in the absence of a food reward, it relates to intermittent reinforcement.
Either “continual reinforcement” and “partial reinforcement” is necessary for both forms of conditioning. Using the rat example of operant conditioning, if the rat is continuously rewarded (constantly reinforced) for pulling the lever, it will continue to do so; but, if the reward stops, the rat will quit pulling the lever almost immediately. It depends. If the reward stops early on when the rat has only a brief history of reinforcement, then it won’t take many repetitions before it realizes that the lever is on a nonreinforcement schedule. However, the longer the history the animal has with being reinforced for a behavior, the more resistant to extinction it becomes. Most falconers have experienced this behavior in the field with birds that have been continuously rewarded for coming to the fist. It doesn’t take long for the hawk to not come to the fist if constant reinforcement ceases completely. On the other hand, the rat will work much harder, even to the point of death, pulling the lever if it is rewarded only on random occasions (partial reinforcement). So, if a falconer continuously rewards his/ her bird with food for coming to the fist in the field, using classical conditioning, operant conditioning. See previous explanation. it will be much less likely to continue coming to the fist when deprived of food, than if it was only rewarded on occasion. If I wasn’t familiar with behavior, this sentence would have completely lost me. A better way to explain this would be that if the falconer wants to increase response due to intermittent reinforcement (which you refer to as “partial”), then the number of calls to the glove that are not reinforced should be gradually introduced; for example, calling a hawk 5 times to the glove without a reinforce is likely to decrease response quickly. Introducing a 9:10 ratio of reinforcement (9 out of every 10 responses are reinforced) and then gradually introducing a 7:10, then a 6:10, etc., should show an INCREASE in response. But, understanding the intricacies of the conditioning process still does not explain why a hawk would return a falconer after an extended period of time in the wild. Yes it does for the reasons I've hit on already and for more that I will elaborate on...
It’s obvious from the behaviors of hawks described by falconers after release that these hawks have long term memory; that’s a given, or they would not recognize the falconer. And, there are a number of stories told by falconers that these hawks remained receptive to their presence even after they refused to come to the fist, which suggests that hawks are not only receptive to conditioning through associative learning, but, once conditioned in this way, the conditioning experience and resultant behaviors seem to be somewhat indelible; however, this is not to suggest that this conditioning cannot be reversed. What is happening here is the hawk has been counterconditioned, and now relates the falconer to positive experiences rather than aversive ones, it has been long-term habituated, and there is a history of positive reinforcement. We know from our rat example that while a hawk might return to the fist after an extended period of time, if this behavior is not reinforced, it will cease to do so; for example, I released a passage RT at the end of its first season in a field that we had hawked fairly consistently. For the first week, I would return every couple of days and feed her. I then extended the time away to three days, at which time, she began to do a number of fly bys when offered food before she would land, but when she did come to the fist, she would keep her wings extended for awhile, which indicated that she was beginning to struggle with being in close proximity of me. But, like Charles’s red-tail, she would continue to hunt with me and the dogs. When I extended the away time away to five days, she was nowhere to be found. This experience is what I would expect—a reversal of her conditioning and a slow transition back to her preconditioned state, in the absence of at least some reinforcement. Of course. First, habituation can wear off quickly without consistent exposure. Second, reinforcers fluctuate in importance depending on associative history, appetitive drive, and the access to other reinforcers. A passage hawk that is released and has access to hunting wild game has other reinforcers available, and they realize they have all day to attempt to gain those reinforcers. The bird will likely also be heavier because of no weight control. I would contend that the more history that a passage hawk has of being restrained on the falconer’s glove, the less likely it will be to fly to the glove in a wild situation.
It appears that if a hawk that continues to return to the fist, after being released for an extended period of time, it does so, because of continued partial reinforcement, not accurate. It would only be partially reinforced if the falconer called the wild hawk down for no reward on one day, and then called it down for a reward on a subsequent day. I think what you’re referring to is that the results of counterconditioning and habituation continue to maintain or strengthen behavior, even when access to other reinforcers are available. and pure, unadulterated conditioning, particularly if the hawk is in good weight. Some might argue that weight has nothing to do with it. They are coming to the fist because they are hungry, regardless of how heavy they might be; in other words, there is a difference between hunger and weight. While this observation is accurate, there still has to be a reason that allows a hawk to even consider coming to the fist, after an extended period of time, particularly if it has been surviving without being fed by a human; otherwise, we could just wave food at any hungry hawk and expect it to come. So, once again, the reason for this return is a direct result of conditioning. I like it. The hawk is coming for an easy, guaranteed meal, that’s it.
Another factor contributing to a passage hawk’s receptivity to coming to the fist after release is length of time in captivity. Based upon some of the foregoing examples, and others, not mentioned, there appears to be a direct correlation between time in captivity and willingness to come to the fist after release—the longer the bird has been in captivity, the longer the period when it will come to the fist without the benefit of reinforcement. That’s due to a longer history of reinforcement. Like the rat pressing the lever, the longer the history of reinforcement, the more resistant to extinction the behavior becomes. As far as tameness goes, the longer the animal is exposed and remains long-term habituated to stimuli, the longer it takes to fade away.
So, what does all of this mean? It means,
1. That passage hawks are intelligent; I don’t think they’re any more intelligent than captive bred birds. They certainly have learned more from their experiences in the wild, so that’s an advantage.
2. That passage hawks are very receptive to conditioning based upon associative learning involving food; All animals are very receptive to conditioning involving food. Food is a primary reinforcer.
3. That this conditioning can be lasting, depending upon time in captivity, and can override any avoidance factor as it pertains to the presence of a familiar person or dog known to the hawk. I wouldn't go as far as saying "any," but yes, for all the reasons mentioned.
4. That hawks have navigational abilities that allow them to find familiar territory, which could an interesting topic for another day.
Dillon Horger
Pennsylvania
