My reading of the literature is that the various stores of domain-specific knowledge may be like narrow abilities with an over-arching broad factor that accounts for the covariance among them being broad Gkn. Ackerman et al have published considerable research demonstrating a broad knowledge factor. In addition, two recent articles in Intelligence present analyses that demonstrate a higher-order broad Gkn (although labeled differently by the the researchers) factor when narrow domain-specific Gkn abilities are present (e.g., knowledge of different domains of current events; technical knowledge, arts knowledge, etc.) --- Two relevant model figures are below (double click on images to enlarge). These are from Reeve (2004) and Hambrick et al. (2008). There are more such studies....esp. by the Ackerman et al. research group.
Mabye the name (broad domain-specific) is a bit confusing....as how can something be "broad" and "specific" at the same time. Suggestions for a better knowledge label? Gk?