In addition to studying complex human reasoning, I have been observing complex instinctive reasoning in animals for many years. Instinctive behavior is often thought of as simple, 
perhaps even mindless. I came to a very different conclusion, that an instinct is basically inherited expert knowledge. Instinctive behavior is often very complex, involving sophisticated concepts, expert knowledge and a lot of decision making.
So I wrote up my observations and hypotheses, in the form of a four year long series of 46 short blog articles. The blog title is Horse cognition and other critters (of which many are discussed). It runs around 30,000 words so there is lots of stuff.
My approach is one I call “robotic analysis”, which means asking what a robot would have to sense, understand, know and decide in order to do what the animal does instinctively.
This does not mean the animal’s behavior is robotic, far from it.
Artificial intelligence robotics tries to model intelligent behavior. For example, developing the self-driving car required a lot of study on precisely what we do when we drive. Knowing the difference between a pedestrian, which can move, and a mailbox which cannot (except under extreme circumstances).
Here are snapshots of a few of the many blog articles, to give you the flavor of the findings:
The amazing cow circle. June 2015
A small herd of cows and their calves is threatened by dogs. The cows do not all face the dogs. Instead they form a defensive ring with each cow facing out. The calves go to the middle of the ring. Consider what is required. The cows have to recognize the threat as calling for a circle. They have to know how to form that circle, which is far from simple, especially since some are facing away from the dogs. The calves have to know to go to the center.
Baby behavior. May 2016
A baby bird hides under my riding mower as I approach, planning to mow. When I sit on the mower it runs out and hides under my truck. This baby knows when to hide, where to hide and how to hide. It also knows when to change hiding places. These are complex behaviors.
Instinctive learning. April 2016
A bird building a nest flies long distances to specific places to get materials. The bird must know what it wants at that point in construction and it must have already learned where the right stuff was.
Not knowing how to trot. October 2013
Two legged critters like us can only walk forward one step at a time, but four legged folks can do it a variety of ways. In horses one of these ways is the trot. However, most Tennessee Walking Houses cannot trot, even though they are as physically capable as other breeds that do trot, which is most of them. Walkers lack this instinctive knowledge.
I began this research many years ago when my wife and I, avid backpackers, camped near an active beaver pond. The beaver pair was still building their dam.
Also Read : Environment-Friendly Tableware is Not Marine Friendly
My first career was designing earth dams for the US flood control program, so the dam interested me.
I was amazed to find that it was much more sophisticated than our dams, which are basically carefully prepared piles of dirt. The beaver dam was a lattice of sticks, supporting a mud and rock wall on the upstream face. Building this complex structure is an equally complex process, including felling trees for the wood materials. I discuss this in the December 2013 article Meet the beaver.
There is also a lot on the robotic analysis research method, as well as on the overall concept of an instinct as an unlearned body of expert knowledge.
When an expert looks at a situation that falls under their domain of expertise, they see and think things that non-experts cannot do. In many cases instinctive behavior in animals is just like that.
Animals live complex lives that we do not understand. Both domestic animal husbandry and wildlife management depend on how well we understand the animals we are caring for or care about. Robotic analysis is a way to understand a lot more.
