Does dog behavior operate according to recognized rules or laws? I would say this. While dog behavior can be described and often predicted in a laboratory, you will find it impossible to predict and control all the behavior of your dog in the real world.
How can that be?
It has long been recognized that animal behavior has an element of spontaneity which has random, Internal and unobservable causes.
Science is about being able to replicate experiments to establish some principles around a given hypothesis. Generally, experiments are not done on individual animals, but instead on populations of animals, and the results are averaged together. Out of those averages, laws are developed to make predictions about future events. However, within experiments there will be random variabilities, which are smoothed out when aggregated into the overall results of the experiments.
Scientists tend to love reductionism, meaning they prefer answers that convert complex phenomena into simpler levels. However, behavior has emergent properties which result from the interactions of simpler elements to derive complex behaviors. People will use the concept that the whole is greater than the sum of the parts. Thus, you can observe the behaviors of an individual ant, but then put that ant in a colony, and a wide range of complex results will happen, such as colony transports; fights with other colonies; nest building; the raising of larvae into adults; and communication between the queen and the workers. For dogs, you can train an individual dog to do this or that action but put that dog into an off-leash park, and an entirely new dynamic will appear and new behaviors and errors in your training will manifest. Now, add in this spontaneity, and it gets even more unpredictable.
Remember the scene in Jurassic Park when Ian Malcolm (Jeff Goldblum) explains chaos theory to Ellie Sattler (Laura Dern)? He was saying pretty much the same thing I am here. Once you take an animal from the laboratory and put it into nature, you can’t always predict the results.
This concept, at the human level, rages between determinism and free will. I have studied this a bit, and science cannot claim that free will doesn’t exist. First because it is untestable. Second, because science has no clear explanation of randomness. While I am not an expert in quantum mechanics, it seems clear that randomness, at the very least, is built into the universe. That also means that if you study the smallest subatomic particles, you can’t then extrapolate what a dog is going to do off leash in the wilderness.
So, are there general predictions that can be made about dog behavior and the results of dog training? Yes. Especially if you have trained a lot of similar dogs in a similar way. However, individual dogs will always present challenges since they aren’t there to follow your rules. We all know that what might work for one dog might not work so well with another dog. What might work in a dog training program might not reflect what that dog will do every time outdoors and off leash. Which is why you play a part in what happens when you are with your dog. A trainer can help you make your dog more predictable, but no one can make your dog do something 100% of the time. That is the risk you take whenever you have a dog. Training is properly described in terms of probabilities for most dogs but should never be interpreted by owners as being an assurance that any given result will happen 100% of the time.
In the end, it is great that scientists and others study animal behavior and training. The more you know about dogs the better you can work with them. However, never get it into your head that you will ever master everything about dogs to the point you can predict what that dog will always do. You can’t do that with people, either. Individuals have a free will, and dogs have something similar going on which is often called spontaneity (though I think it is a lot more than that).