The issue of in what way distributions can be different is discussed in #411. I think we need to resolve that issue once and for all.
But apart from that, I think that should be no restrictions to what can be cited, so the object of a citation could be distribution.

Related to what we provided for dataset citation: #61

Discussion in Sprint meeting was that

• the lack of domain constraints on properties needed for citation (e.g. dct:creator, dct:publisher etc) means that they could be used on a dcat:Distribution if required
• we are reluctant to copy more of the properties associated with a dcat:Dataset down to the dcat:Distribution class, as it muddies the model. We do want to preserve the core distinction between datasets and distributions
• we generally expect citations to be to datasets. If only a specific distribution is intended then that should be indicated with additional narrative alongside the formal citation

The consensus was not to change the model for dcat:Distribution.

https://www.w3.org/2019/02/05-dxwgdcat-minutes.html