Who drew the line between abstract/concrete and general/specific in the philosophy of language? Are there any comprehensive resources on the early history of their development specifically with respect to language?
Like many terms in philosophy, 'abstract/concrete', 'general/specific' derive from the Latin used by the schoolmen in the middle ages.
The adjective concretum means hard, solid, material, whereas abstractum literally means drawn out from, i.e. abstracted. It is in some ways similar to the distinction between material and immaterial. William of Ockham discussed it chapters 5-9 of book I of the Summa Logicae, but as he is a nominalist he argues that the distinction is linguistic only. 'Humanity" is a verbal modification of 'human', and "Socrates has humanity" means no more than "Socrates is human".
'General' comes from the Latin 'genus' meaning 'race' or 'kind', which is a translation of Aristotle's Greek word genos. 'Specific' comes from 'species', meaning outward appearance, shape or figure, and is a translation of Aristotle's eidos.
"Are there any comprehensive resources on the early history of their development specifically with respect to language?"
Difficult - you could try looking at Joyce's Principles of Logic, which is a manual of traditional (not modern) logic. Mill's System of Logic covers the abstract/concrete distinction in Bk I c ii sec.4, and see c.vii of the same book, on the 'predicables', the traditional part where the genus/species/differentia &c division is covered.
I don't know how early you want to go, but the most celebrated pronouncements on this topic in modern times are probably Wittgenstein's. One is "For philosophical problems arise when language goes on holiday" (Philosophical Investigations 38, 1953). I understand that to be an implied critique of philosophical abstraction, when language is divorced from its workaday use as a practical tool, which is the role from which it derives its true meaning and vitality as a "form of life".
As for your parallel binary, general/specific, see the Stanford article:
"It is here [PI 65] that Wittgenstein's rejection of general explanations, and definitions based on sufficient and necessary conditions, is best pronounced. Instead of these symptoms of the philosopher's “craving for generality”, he points to ‘family resemblance’ as the more suitable analogy for the means of connecting particular uses of the same word." (Stanford Encyclopedia of Philosophy, Wittgenstein, 3.4)