I just watched this really excellent talk by Tim Ingold on Anthropology beyond Humanity. I found it to be fascinating and engagingly delivered but argumentatively flawed. First, Ingold claims that Latour claims that there is a radical disjuncture between humans and non-humans with regard to the socialisation of objects. That huge diagram behind him disproves that immediately! Non-human primates are included on the same diagram as humans. So, there's a continuum, not a bifurcation! I think Ingold has a point inasmuch as Latour doesn't adequately distinguish between different kinds of objects and how they differently socialise humans and allow humans to socialise but his point is overstated.
Secondly, he gets Latour's 'collective' all wrong. The collective, a I understand it, is a political concept. To be within or without the collective is a political means/ends distinction. When Latour says that some non-humans must be 'outside' the collective this doesn't mean that they aren't part of human life as such - precisely the opposite, in fact. It is because humans are now entangled with the entire planet that we need to decide what is inside and what outside. Historically, during modernity, nearly everything was treated as a means to humans' ends (masters and possessors of nature, etc.) - that is no longer tenable because of Gaia, etc. However, we can't then make all things ends, either. We wouldn't even be able to eat. So, we have to decide. It is because we are entangled with all earthly existence, so the argument goes, that there must be an 'outside' to the collective. The reason why non-humans need 'representatives' is because their case has to be argued politically, which is most definitely something that cannot be done by non-humans themselves (Ingold argues that non-humans can 'speak for themselves' and I suppose they can but in a very limited way with respect to politics).
In all of this Ingold conflates different meanings of 'social' - there's the simple sense of one thing affecting and being associated with another (in which case humans are social with the entire planet, as Latour has argued again and again - in the way Ingold describes it at some points we might have to say that we are social with the entire universe) then there's social in the sense of fixing or unfixing social structures or social habits, in which case humans are most definitely social with non-human things but in a more limited sense (and so it goes for non-human sociality, too). I think Ingold is right to point out that non-humans are just as involved with other kinds of non-humans as humans are and that this could be called 'social' in some senses of that word but he's wrong, I think, to suggest that Latour refuses these entanglements. It's simply not a matter of 'non-humans amongst themselves' versus 'humans plus non-humans.' I think that's a false reading. Humans enrol non-human things in order to fix, reify and, if you like, mineralise their social relations (and thus make them extend over time and space) - humans do this at a far greater intensity than e.g. primates. I see little to object to here. Where Latour goes wrong, I think, is in arguing that all technological artefacts constitute this broader sociality. Ingold points out that often tools are made from local materials and immediately discarded. That is right. The kind of artefacts that mineralise social relations are e.g. the colourful feathers made into a ceremonial headdress that reminds everybody of who the chief is (a somewhat ethnocentric example but, I hope, a fairly innocent one). Fashioning a stick into digging implement and then immediately discarding it doesn't do much for social relations at all. So, Latour's argument is a bit simplistic but I'm not at all convinced that it's wrong in the way that Ingold is arguing here.
Lastly, the idea that we must think of species in terms of ongoing practices of speciation is a fine one although pretty much compatible, I think, with Latour's arguments in general.
Overall, Ingold seems very determined to damn and blast Latour and goodness knows that I do a bit of that myself but his understanding is a bit shallow and some of his comments in the Q&A at the end are simply prejudicial. He doesn't like the fact that Latour uses a speedbump as an example of non-humans because, for Ingold, humans are much more rich and diverse than that and they can be learned from as agents in their own right. But when Latour says 'follow the actors themselves' he isn't just talking about humans! Ingold makes it sound like Latour is only interested in humans and non-humans are simply instrumental and secondary, which is precisely the opposite critique that the vast majority of people have made of Latour, which tells a story in itself. Latour clearly irks Ingold deeply (a common enough reaction, for sure) but unfortunately his ire is largely misdirected.