Since the late nineteenth century it has been widely acknowledged that there is a significant proportion of vocabulary of uncertain origin that is not from the Indo-European inherited lexical stock, but apparently borrowed from the languages already spoken in the Aegean before the introduction of the linguistic ancestor of Greek into the region. Some potential sources of these loanwords are known (e.g., the language of Linear A, Eteocretan, and other non-Greek languages concretely attested in the epigraphic record) but remain undeciphered. Still others may have been completely lost. Many theories have been proposed to account for these loanwords and none has met with scholarly consensus. In this paper I will review the arguments that suggest a plurality of linguistic sources for these loanwords (contrary to recent views, e.g., Beekes 2010, 2014) and propose a preliminary framework as an initial step towards identifying and delimiting layers of prehistoric lexical borrowings in Greek.