Misapplications of the archaeological culture (hereafter abbreviated AC) concept abound, but the concept itself is straightforward. As she does with so many things, Pat Galloway (1995:28) explains it well:
Named (archaeological) cultures, or “phases” thereof, are in fact sets of artifact types and other physical remains (dwelling structures, settlement patterns, burial practices, and so forth) that occur together consistently over a definite spatial extent and – as far as can be determined – over a definite continuous time period.
While the validity of individual named ACs is often called into question (e.g., Hart and Brumbach 2003), the validity of the AC concept in and of itself is generally accepted.
The AC concept sometimes gets a bad name, however, because of the way in which it was/is used within the culture-historical approach to archaeological research. The culture-historical approach seeks to correlate ACs with documented or reconstructed social groups. While 1:1 mappings of ACs to social groups continue to be made by researchers, such as Kurgan:Proto-Indo-European or Fort Ancient:Shawnee, archaeological theory has come a long way since the heyday of the culture-historical approach in 1930s/40s. This means, in practice, that contemporary archaeologists proposing such correlations are held to a much higher burden of proof by fellow archaeologists than would have been the case seven decades ago.
Galloway, Patricia Kay. 1995. Choctaw genesis, 1500–1700. Lincoln: University of Nebraska Press.
Hart, John P., and Hetty Jo Brumbach. 2003. The death of Owasco. American Antiquity 68 (4): 737–52. doi:10.2307/3557070.