An increasing number of studies have shown a variety of animals experience some emotions. The BBC has been examining some of this in "the Wonder of Dogs" documentary series. Many studies have examined the dog-owner bond showing they react similar to children when their owner leaves, and even that your dog will 'catch' your yawn. Emotion is not unique to dogs many animals show responses suggesting they experience emotion (just google "animal emotion" and you'll find tons of references). Recent work with MRIs have even show similarities between human and dog brain functions.
As with many things I am surprised by other people's surprise at this. The only explanation I can come up with for why people are shocked by this is that people think emotions are big complicated things because we humans struggle to understand them. Therefore you must need a big complicated brain to produce these powerful big complicated things we call emotions.
But from my perspective emotions are not big complicated things. True we struggle to understand them. But that is because we try to apply our higher brain functions of logic and reason to them. To explain why this is silly I invoke my past discussion of the inefficiency of evolution.
Remember, evolution is not good at finding 'the best' solution to a problem. Evolution tends to find the simplest solution which is 'good enough' to survive. When we consider behaviour it only matters whether you do or don't do something, it doesn't make any difference whether you understand why you do it.
The other important thing to realize about behaviour is that for many many animals there are and extremely large number of possible situations they could be in. Any programmer or roboticist will know first hand how hard and time-consuming it is to write a program able to deal with many possible situations. A whole field of machine learning has been developed to help with this. Evolution is just too inefficient to hard-program behaviour for every situation experienced by even simple animals.
Learning is ubiquitous in the animal world. Even some insects, such as bees and wasps, have been shown to be able to learn. Learning has been extensively studied due to the need to train dogs and horses (among others) to work with us. Learning comes down to simply reward and punishment. We experience reward as pleasure as pleasure and punishment as pain. Both pleasure and pain can have physical, emotional, and psychological dimensions, which are intertwined in the brain. We should expect any animal which can learn to be able to feel pleasure and pain. Likewise any social animals should be able to feel pain of rejection and pleasure in friendship.
Let's take a specific example: Most birds rely of flight to escape predators. Developing flight takes a long time but there are physical limits on how big an egg can be. This means their offspring must be flightless and vulnerable for a long time. It the parents take care of the offspring they are more likely to survive increasing the evolutionary fitness of the parents. There are a few routes open evolution could follow. 1) hard program the parents to regurgitate food whenever the offspring are nearby and to fly away squawking if a predator approaches the nest (distraction). 2) program the parent feel pleasure when their chicks are full, warm and content in the nest (ie. caring). Make the parent feel aggressive and fearful when a predator approaches the nest (ie. protective). 3) program the parent to understand that their children share genes with them so helping them survive will increase the number of copies of their genes in the population and that is the sole purpose of their existence. Clearly #3 is excessively complicated (most humans do not accept it), #1 may or may not be harder to evolve than #2 but it should be obvious it will be less effective because it doesn't provide the flexibility to use different strategies depending on what the predator is (sometimes distraction might be more effective than attack or vice versa) it also cannot account for differences in terrain or weather etc... A more nuanced program would be much more difficult to evolve. So in most cases we expect to seen option #2.
Intuitive emotions are more powerful and simpler solutions to getting organisms to behave in ways which benefit their evolutionary fitness than solutions which rely of organisms understanding why they should behave that way or than solutions which hard program specific behaviour patterns. So we should expect them to be ancient and present in many organisms.
