For Latour all things from people, trees, minerals, all the way to atomic phenomena are able to be seen as actors and actants, with no actor or actant being reducible or irreducible to anything else. A tree cannot ultimately be reduced without arbitrariness to bark, wood and chlorophyll systems any more than it can be to its molecular or atomic structure. Any whole is from this perspective always at least conceptionally 'less' than its parts. There is a indissociable slippage between macro and micro multiplicities which seems to me to be very similar Deleuze and Guattari's notions of molar and molecular. The notion of alliances between actors is also very important for Latour, which determines the weight of an actors existence in the world, with an entity existing with more and more reality the more alliances it enters into.
Actors from this perspective exist relationally, and there exists only 'vectors of being', with both events and objects considered to exist, but not necessarily on the same footing: “An entity gains in reality if it is associated with many others that are viewed as collaborating with it. It loses reality if, on the contrary, it has to shed associates or collaborators—animate and/or inanimate.” (Latour, 257))
Within this framework Latour also retains the notion of essence, but construes the term in a curious way: we are able to know they exist, and have a general or vague intimation of what they are, but it is an essence which is so weak that it is unable to be definitively captured at any level. It is said that Latour is criticised both by realists for being too constructivist, and by constructivists for being too realist, and that his position sits somewhere between the two.
Why retain the notion of essence within this formulation, is it not a redundency? Does this contribute anything to the 'hot-potato game' (so to speak) of irreducibility between micro and macro levels of phenomena ?