To be a little obtuse and deliberately try and start a little fire on this, you could always come back and argue that the entire idea of imagination is, in itself, an illusion. Like consciousness, it's a retrospectively created by-product of what are, in actuality, localised and largely thoughtless instinctual responses. We like, we do. It hurts, we don't. Big fat endorphine rush, intellectualise later. We didn't ever imagine Cathy and Heathcliff, but we imagine we did. Just like 'we' don't exist, only the idea of 'we' that 'we' find useful/rewarding in that moment. That's not just solipsism (it's not actually solipsism at all) but increasingly there are ideas floating out of cognitive science that this notion of cohesive, pervasive cognition, emotion, type is actually largely illusory.
So when we talk about imagination as a user requirement, or preference, or even about the artists' conscious decisions about accessibility, how does this depressing/liberating (delete according to personal taste) idea affect how we think about it?
Or, cast in another light, are we simply peddling the same old opiate just wrapped in another set of predelictions? At what point do we understand whether the 'deeper', 'truer', 'higher' content or reaction or imagination we strive for (and I think most people on this forum tend to consider their work as striving, and not in an elitist, arrogant way, but in a kind of 'why climb the mountain' way, if that makes sense) as being essentially exactly the same kind of mental ketamine as Just Cause 2 or Crysis, but one that feels different to those who like it because their liking of it is predicated upon them feeling that it feels different.
I'm stopping now because I think I've confused myself. Does any of that make any sense?
The meta-philosophical point of interest in debates like this is that the only way we may deal with the world in a 'reality' perception is by forcibly not explaining things - we are made of atoms, it is inevitable that at some point of granulation 'we' do not have anything coherent to define. In that interest we ought to define imagination.
If we take the approach of Damasio, we can see the core of the brain, the 'primitive' part, so to speak, as the receptor for sensations from the body. Skipping the part which interprets these signals, there is apart of the brain which receives signals, which we may correlate with feeling something. I.e., a certain stimulus is warm, another cold. This is not imagination, of course.
Damasio adds the expanded consciousness, which is capable to synthesizing false data (in the sense that it is not derived from actual nerve data) and feeding that to the core - producing sensations without a bodily cause strong enough to seem real. This synthesis is what I believe has caused higher lifeforms to thrive as it allows us to make countless permutations and raise expectations. At the same time, synthetic sensations seem so real to us because the control mechanism is the mechanism which also created the patterns for these synthetic sensation - like Hinton's deep belief networks re-creating a glyph.
This is how I believe we can define imagination: first of all, there is this capacity with factors such as strength, endurance and quality. Quality would be the capacity of fooling one's own control mechanism. There are things which I can imagine so vividly it may as well be real (in certain moods) and there are things which hold no bearing to real life, especially things which I have not physically experienced and have no 'original' data of.
In this sense, art (or any medium) transfers symbolic data which one part of us translates into sensations which may be fed to the core brain. It can be seen as an aid. If I do not have the translation table I cannot read the glyphs. If we imagine we are noblemen from the 17th century donning the right clothes may be of aid to us because it helps complete our false data. Vanitas can be an aid as to imagine different things.
But I think larger beliefs are also run through the primitive brain - it knows no scale so we may imagine 10 billion people or a 100 with the same mechanism. There is more a
bandwidth problem in a sense. But different experiences do exist, it cannot all be generalized to ketamine. There are things which are great thrills (like a vivid firefight) but leave me empty after finishing them. A good book, however, becomes integrated as part of my thinking.
The mental integration is an important part of Ayn Rand's philosophy of art. She specifically applies it to music in the sense that we try to generalize what we hear and in a manner which is in a sense trying to mimic the more fuzzy data with mathematical constructs. The more apt the brain is, the more layers of constructs you would want to fulfil at once. She is highly dismissive of non-tonal music exactly because it dulls the brain into receiving
impulses rather than learning it to generalise data.
In this I follow her and applying this, one may say some forms of imagination are for the primitive brain (base sexual fantasy, for instance), others are vivid enough to be integrated with the real brain. An imagination is so vivid it is stored as a real memory. The difference between Crysis and Wuthering Heights to me basically exists in that while both have a 'in the moment' sensation it is only Wuthering Heights which has larger, more complex lines which engage with (and add to) wider parts of my brain and which has a rhythm I can grow accustomed to, takes my mind to the cold north and confronts me with questions of honour and human nature, adding to the concepts in my mind as well as linking them in new ways. For instance, I may without any structured thought tell you whether I think someone 'a Heathcliff' because I have neurons dedicated to providing data as to recognising Heathcliff. Crysis does not have any this sort in my mind, it operates far more closely to my primitive brain, so I may recognise the outline of some of its adversaries.
When I remember Heathcliff I remember many facets and I notice when I talk of good books to friends the experience never quite comes out the same way, because each of them is more prone to a 'subset' of my Heathcliff generalisation, which is a memory of a person which functions in the same manner as a memory of a real person. In this sense the quality of my Heathcliff consists of how well my imagination worked on the Heathcliff symbolic data.
Speaking for myself, my 'mountain' is allow the medium of games to create these longer, generalisable experiences. I have not had this much with games and plenteous with books. In my personal opinion, I think the medium of literature for many types of experiences to be far superior to film and games. But this is more a problem with the manner in which we make films and games, I am sure.