I started Niels Peter Lemche's The Canaanites and Their Land. On page 6, Lemche states the thesis of this book:
"The main thesis of this book is that the 'Canaanites' of the Old Testament are not a real nation but an imagined nation placed in opposition to the Israelites. The Canaanites of the Old Testament are the 'bad guys', hated by God and the world, doomed, and to be replaced by the Israelites." As I said in my last post about Lemche's Israelites in History and Tradition, Lemche holds that this took place in Israel's post-exilic period, when a group within Israel demonized the inhabitants of the land to justify Israel's possession of it.
But aren't there references to Canaan in the ancient Near East? The answer is "yes," and Lemche discusses those references. Here are his conclusions:
"...Canaan was never a well defined and identifiable territory and...the term 'Canaanite' meant little to the people that lived in the Southern Levant in antiquity" (page 6).
"To the scribe of ancient West Asia 'Canaanite' always designated a person who did not belong to the scribe's own society or state, while Canaan was considered to be a country different from his own. In this way Abi-Milku [of Tyre] was able to include Ugarit in his Canaan, although the citizens of ancient Ugarit never considered themselves to be Canaanites. It might well have been the case that conversely, in the eyes of the Ugaritic administration, Abi-Milku's city was situated in Canaan, although Abi-Milku himself never says so." (Page 52)
And so, for Lemche, "Canaanite" in the ancient Near East usually referred to an outsider, and my impression is that the outsider is often West Asian. But there isn't consistency within ancient Near Eastern references to Canaan about the boundaries of the land. In an El-Amarna letter from the second millennium B.C.E., Abi-Milku of Tyre notes that the Pharaoh asked him to report on events in Canaan, and he proceeds to discuss what's occurring in Ugarit. For Abi-Milku of Tyre, "Canaan embraced [the area] from Damascus in the south to the Hittite border in the north" (page 30), which is in eastern Cilicia (at the bottom of modern-day Turkey) and inner Syria. Anson Rainey has a different interpretation of the Abi-Milku passage---that it means that the Pharaoh is wanting Abi-Milku to report on events in Canaan, meaning that Tyre is a part of Canaan. But Lemche seems to argue that Rainey has a vested interest in distinguishing Ugarit (in Syria) from Canaan, for Rainey wrote an article on a text in which "a foreign merchant visiting Ugarit is described as being a 'Canaanite'" (page 31). For Lemche, the situation appears to be that Abi-Milku in Tyre defined Ugarit as Canaan, whereas Ugarit viewed a foreign merchant as a Canaanite, meaning that Ugarit lacked a Canaanite self-identification.
A Hittite text from the second millennium, however, does not treat Canaan as either Tyre or Ugarit, forit distinguishes Canaan from those areas, even as it "combines Canaan with the coastal states along the Mediterranean, indicating that according to his understanding of Canaan this country was also situated along the Mediterranean coast" (page 42). And, indeed, Lemche says that the Egyptians regarded Gaza as the center of Canaan, for, "since the early days of their sway over Asia, [the Egyptians] used Gaza as their most important base in Palestine" (page 47). A second millennium letter from the king of Babylon to Pharaoh, however, refers to a city in Canaan that may be in Galilee, in the north of what today is Israel (page 33). And Lemche says that "Canaanite" could also function as a sociological designation, for a second millennium Mari text places "Canaanite" in parallel with an Akkadian word that is often used for "brigand"---and so the text may be saying that the "Canaanites" are outlaws, or even "highwaymen of foreign origin" (which is Lemche's quote of Manfred Weippert on page 28).
Regarding the first millennium B.C.E., Lemche says that Canaan at that time goes unmentioned in Assyrian, Babylonian, Egyptian, and Syrian documents, but that there is evidence that some Hellenists equated Canaan with Phoenicia.
My impression is that Lemche is saying that the Canaanites were not an ethnic group in which one was "in" or "out," and knew that one was "in" and "out." People used "Canaanite" for outsiders, usually Asiatics, but there was no group regarding itself as "Canaanite." Plus, in the texts, "Canaan's" location is rather vague.
But wasn't it the case that the Hebrew Bible defined the extent of Canaan in terms of the boundaries of the Egyptian dominance over that region, after Egypt's treaty with the Hittites in the thirteenth century B.C.E. (see here)? Lemche says that the Hittite and Egyptian versions of the treaty do not specify "the precise location of the border between the empires" (page 68), which (for Lemche) may mean that we cannot look to those treaties to learn the definition of "Canaan" in the ancient world.
I'll say a quick word about the Hebrew Bible's lists of Canaanite nations, according to Lemche. Lemche says that the Hebrew Bible is generally precise about the location of Canaan, but that it lists nations, such as the Amorites (whom Lemche locates in Syria) and the Hittites, who were not in Canaan. For Lemche, the biblical authors are tossing in some names that they know from tradition, in order to put non-Israelites into the land for their story (pages 91, 100). These non-Israelites would represent the non-Jewish inhabitants of Palestine in Israel's post-exilic period.