Emotions are recognized nowadays as having an important role to play in our reasoning, rational or irrational. Studying consciousness in an attempt to develop systematic models and conscious machines puts us face to face with this fact: emotions play a strong basis for our psychic, reasoning, personality, and conscious. But what are emotions? Where do they reside? How do they relate to consciousness and self-image? Let us hold to these questions for a second and explore another area of brain/mind modeling.
In hypnosis and Freudian psychology we are often faced with the concept of subconscious. It is the hidden psyche which we need a process of hypnosis, dream analysis, and other mechanisms to reach, to bring out, and to affect and change. Without these techniques of accessing subconsciousness it stays hidden but very influential. We almost have to be unconscious to be in subconscious access mode. And change in subconscious is very slow and difficult to achieve. Now let us hold these thoughts while exploring another area of brain/mind modeling.
This time, let us look at artificial intelligence and neuroscience attempts in modeling brain/mind. There is a long standing argument between matter (i.e. brain) and content (i.e. mind). I am not going to explore this argument. However, let us explore the neural modeling of brain/mind, which comes from the connectionist school which believes matter and content is one thing and resides in the brain cells connections.
In the neural models, we have input layer of neurons (i.e. nodes) and output layer. In between we may have nothing, i.e. direct connection between input and output, or we may have hidden layers. The input layer could map to perceptual senses whilst the output layer often maps to behavioral actions. There have been attempts to model emotions and feelings using neural nets. Thinking about this concept in which feelings are entered through senses (i.e. input layer), emotions calculated in the hidden layers, and behavior produced as output layer values. The input and output layers struck me to be very similar to conscious observation of the environment and conscious actions within the environment; but unconscious or subconscious analysis of perceived information may relay on emotions which could be extended to what I may call perceptual emotions.
Perceptual emotions would include sets of emotions related to self-image (i.e. what do we think of ourselves), to others who are consciously being observed (i.e. what do we think of other people and objects in the environment), and so on. These perceptual emotions reside in the hidden layers of our neural composition forming the basis of our subconscious and deriving our cognitive processes such as memorization and thinking. These perceptual emotions often manifest themselves in the form of beliefs, which greatly influence our reasoning and actions and very difficult to change. In fact changing one's beliefs may mount to changing their persona and individualistic print.
Can we conclude from this that consciousness is the pure sensing of surrounding and behavior, including sensory errors, while subconsciousness is the individualistic perceptual emotional model of that sensing with its all memorization implications? If we agree with this conclusion, one may derive that animals must have conscious and machines could have conscious but neither of them necessarily have or could have subconscious. Personally, I agree with the first part of the conclusion but doubt the second part especially with living entities. My guts feeling is that they must have some form of subconscious.