I would tentatively say yes. Now, everything else I say is predicated on the idea that we will first have to be able to qualify/quantify emotion so we can look at a 'feeling' (My guess is some form of advanced FMRI) and see if it's a novel feeling, a subset of some other kind of feeling, or an entirely new form of 'feeling.' To give an example, if I create a well researched, thorough, fair and easily digestible post or comment on Hubksi/Reddit/etc, and that post doesn't do well, or is downvoted, ignored, one could say that I'm disappointed, but not in the same way that I would be if I had been stood up for a date, or was going to be late to an important event. This I think, is a failing of language, not of emotion. It's like the idea that Inuit people have hundreds of words for snow and ice, because they see distinctions between those types of snow and ice that a non-native doesn't have language for. So, as we delve into the neuroscience aspect of 'feeling' we should also explore, and thoroughly, the language of 'feeling' so that [The Sadness at the death of a parent] is differentiated linguistically from [The Sadness at the end of a teary movie] English, to give it some credit tries to distinguish 'sadness' from 'anguish' and 'sorrow,' to give examples, but because of the synonymous nature of English, they become symbolically equivalent. If you're 'sad' you can also be described as 'anguished' or 'sorrowful.'