Lomax hasn’t replied, so I’ll have an attempt at this- feel free to fix it.
I think for Davidson, although there seems acknowledgement that essence etc. is problematic, as maybe Quinean answers to the problem may also be, there is attempt to “go along” with notions of say, sentence truth, roughly, allowing for such things as metaphors, and adding notions of “understanding”. From this, I think Davidson allows for understood meanings to change which may involve some sort of rough overlap or encompassing other terms? So perhaps the above contrarily, e.g., might suggest that usage of terms as metaphors may give definitive meaning, or terms used while not understood, e.g. mistaken usage, or a sentence with no meaning that may be true but lack any meaning, may give definitive meaning.
Derrida also seems to acknowledge problems with intentionality, essence, etc., and differently from Davidson’s “going along” with such notions, holds these “under erasure” while in use, as if involved in such discourse from a more removed or external position, but perhaps allowing that even from such a position, use of such sort of logic may still disrupt or corrupt outcomes. (Incidentally, this is something I’ve wondered about in some arguments in various threads like this. Questioning the logic of arguments upholding notions or theories if from differing positions without alternative, may seem more or mostly about questioning or denying, the logic of the arguments against, themselves- although from a Derridean perspective, logic users including me may also be susceptible to some extent to own petard hoisting too, I think. I’m not sure how much if any immunity notions like “under erasure” allow.)
With iterability, Derrida adds citations and I think, an amount of alterity. This could include citing terms in new contexts with seemingly very different meaning, if any. Or an obvious example would be a married man playing the role of a bachelor in a performance. More realistically, much of the history of Christian marriage was vulnerable to a possibility of annulment, retrospective denial of life lived during marriage, yet, e.g., legitimacy of off-spring conceived in the non-marriage duration. Related meanings of bachelor and the now effectively obsolete spinster have changed a great deal from usage around the 1300s, changing with changing economic systems, class-ist, gender-ist, age-ist, nuclear family-ist, notions, etc. I think, from Derrida is a notion that each reception of terms brings some amount of different meaning, yet there seems to be some on-going similarity that is iterable despite differences. Iterability might involve some more subjective interpretation linking similarities, while differences may perhaps be less subjective, more external, distanced from essence and motivational notions- as some measure of meaning.
Regarding a question I had about function, I still haven’t thought it through and could be making quite a large mistake, and wonder about help doing so! It may be just an ad hoc measure, perhaps like Davidson’s understanding, or Derrida’s under erasure, perhaps also considered under erasure. Related to this, I think it is interesting that here there is talk of usage of language as though it is a functional artefact. Derrida suggests we are like texts, language emanating, experiences inscribed upon us, and perhaps more, inscribed memories, DNA, etc., may further this notion. I’ll list some reasons why I’m wondering about this. One is mereological concerns or about involving fewer assumptions, or at least, some restriction to a set of assumptions, so possibly easier to investigate. Others seem to consider these, e.g. from another link provided somewhere, Searle,
https://www.youtube.com/watch?v=Rbthpbs6PJI from around 40:00 to 45:00 suggests similar concerns, but perhaps includes some presumptions. Along with this is reasoning that the paper I linked suggests that interpretations involving function may not be easily reduced. Another reason is that it may trace language linkage better and allow any notions of synthesis to be clearer. Another is a suggestion that functionality may be less problematic in terms of twin world scenarios (or otherwise more so, requiring inclusion of large system of functional relationships).
(Should have written Francois, before.)