If you ask me what kind of thing artworks are, I would be tempted to say that they are all a form of performance. An artwork is not an object, it's not a visual field, it's not material thing you gaze at: it's an artist's performance into an artistic medium.
What the heck? Well, think of it this way: given two indistinguishable physical objects, one being a perfect accident of nature, and the other being Michaelangelo's Moses, would you interpret and appreciate both the same way? Not really, insofar as one object demonstrates someone's actions, someone's WORK. Not just a beautiful surface or object, but the "genius" of someone, talent, meaningfulness, intentionality.
The above does not constitute in itself a definition of art, since it's an ontology (and I gleaned it from a book called Art as Performance by McGill philosophy teacher David Davies. It's quite a read.). One could say for example that my ontology also applies to cooking: we're not just appreciating the taste, the flavours of the food, but also the work in the medium of food, the talent, the genius, etc of the cook.
So a proper definition of art would require necessary and sufficient conditions. The above ontology would be a necessary, but not a sufficient condition. And to be frank, I don't think I have a sufficient condition(s). So I would be tempted to inch towards what is called "cluster definitions" of art. In other words, to be a work of art, you do not have to possess properties A, B, C, and D, but you could get by with A, B and D or B, C, and D. You call that jointly sufficient conditions.
I would probably argue that property B corresponds to "being a performance in an artistic medium" as explained above, so I would not consider a definition of art bereft of property B, but if an artwork gave me B + 2 other properties from A,B,C,D, I am satisfied.
At least, that's where I am so far.