6.1 A. Conclusions and Caveats
“These are exciting times for those involved in research, development, and the use of intelligence test batteries” (McGrew, 1997, p. 172).  This 1997 statement still rings true today.  Central to this excitement has been the recognition and adoption, both within the theoretical and applied fields of intelligence research and intelligence testing, of the Cattell-Horn-Carroll (CHC) theory of human cognitive abilities (or some slight variation) as the definitive psychometric theory upon which to construct a working taxonomy of cognitive differential psychology.  Echoing Horn (1998) and Jensen’s (2004) comparisons to the first presentation of “Mendelyev’s periodic table of elements in chemistry” and Hans von Bülow’s “there it is!” declaration upon reading the score of Wagner’s Die Meistersinger, the order brought to the study, measurement, and assessment of human cognitive abilities by Carroll’s (1993) synthesis, a synthesis built on the shoulders of a crowd of psychometric giants (Horn and Jensen included), has finally provided both intelligence scholars and practitioners the first empirically-based consensus Rosetta stone from which to organize research and practice.  This is truly a marvelous development.
Human intelligence is clearly multidimensional.  The past decade has witnessed the accumulation of evidence that supports the broad strokes of the hierarchical multi-ability CHC theory of human cognitive abilities.  This new evidence, often derived from studies that gathered data with a wide breadth of ability indicators in large nationally representative samples, validates the inclusion of the broad (stratum II) abilities of Gf, Gc, Gq, Grw, Glr, Gsm, Gv, Ga, Gs, and Gt in the CHC taxonomy.  In addition, past and recent research suggests the need to attend to, and possibly incorporate, knowledge of additional broad domains (Gps, Gp, Gk, Gh, Gkn, and Go) in future research, measurement, and assessment activities.  Although not reviewed here in this chapter, it is also important to note that CHC theory is not based solely on factor analytic evidence.  Developmental, outcome-criterion prediction, heritability, and neurocognitive evidence add to the network of validity evidence in support of contemporary CHC theory (see Horn & Noll, 1997).
Consistent with Carroll’s (1994) self-critique and recommendations for future research, it is important to recognize that the CHC framework is “an open- ended empirical theory to which future tests of as yet unmeasured or unknown abilities could possibly result in additional factors at one or more levels in Carroll’s hierarchy” (Jensen, 2004, p. 5).  The cognitive speed ability research reviewed here serves as a cautionary tale that should ward off the tendency to succumb to a “premature hardening of the CHC categories.”  The importance of this caveat was clearly demonstrated this past decade vis-à-vis the structural research on the domain of cognitive mental speed, where research now suggests a domain characterized by a complex hierarchical structure with a possible g-speed factor at the same stratum level as psychometric g.  In this case, the CHC taxonomy has been used as the open-ended framework described by Jensen (2004) and as Carroll’s (1994) intended “guide and reference for future researchers” (p. 22). 
The revisions, additions, and extensions to the CHC taxonomy suggested in this chapter are based on a reasoned review and evaluation of research (primarily factor analytic) spanning the last decade.  It is hoped that the proposed CHC theory modifications proposed here enhance the “search for the holy grail” of human cognitive ability taxonomies, at least by providing a minor positive iteration towards convergence on a more plausible model.  However, the proposed CHC taxonomic enhancements summarized here require additional research and replication.  Reanalysis of Carroll’s 460+ datasets with contemporary procedures (viz., confirmatory factor analysis-CFA), combined with both CFA and Carroll EFA-based exploratory procedures of post- Carroll (1993) datasets, will help elucidate the validity of current and future proposed revisions of the CHC taxonomy. 
Finally, although additional cautions and limitations could be enumerated, the seductive powers of a neat and hierarchically organized structural diagram of cognitive abilities must be resisted.  Any theory that is derived primarily from a “rectilinear system of factors is…not of a form that well describes natural phenomena” (Horn & Noll, 1997, p. 84).  By extension, assessment professionals must humbly recognize the inherent artificial nature of assessment tools built upon linear mathematical models.  As stated by MacCallum (2003):
it is abundantly clear that psychological researcher’s make extensive use of mathematical models across almost all domains of research…It is safe to say that these models all have one thing in common: The are all wrong. Simply put, our models are implausible if taken as exact or literal representations of real world phenomena. They cannot capture the complexities of the real world which they purport to represent. At best, they can provide an approximation of the real world that has some substantive meaning and some utility (p.114-115)

[Note. See Carroll (1994, 1997) and Horn and Noll (1997) for excellent self-criticisms of the CHC theory by the primary contemporary theory architects. ]