In literature I generally see references to people much more than to source. When faced with ambiguity, I prefer choosing what sounds more correct grammatically, even if the difference is slight.
In the example above, it should be 'X and Y (2011) describe', because the verb 'describe' would apply to an animate subject (person) rather than an inanimate subject (source). Similarly, demonstrate, suggest, propose, hypothesise would be apt for people.
The source would be better referred to when using 'inanimate verbs' like contains, comprises, constitutes etc.
Examples:
X and Y (2011) propose that trees be considered sapient.
X and Y (2011) contains compelling evidence for trees to be considered sapient.