hypothetical "rational agents of construction" (304), the agreement
that justifies the selection of a principle is not merely the
agreement among these agents. Rather, the choosers are components of
a procedure whose features collectively model the faculty of practical
reason. As discussed above, the choosers represent the rational, an
element of practical reason which the procedure subordinates to the
reasonable (represented by the formal conditions on principles
adopted, the veil of ignorance, the focus on background justice, and
the symmetrical situation of the parties). A principle is justified,
according to Rawls's account of Kantian constructivism, only if actual
moral reasoners, adopting the original position as a perspective,
judge that the principle should be accepted. The original position is
thus merely "an expository device" (TJ 19) that is designed to "focus
their moral sensibilities" (TJ 46) of actual reasoners in their
deliberations. Kantian constructivism, like Kant's categorical
imperative procedure, thus requires the agreement of actual--not
hypothetical--persons in order to justify moral judgments. Krasnoff's
argument mistakes an element of the model (the agreement of the
hypothetical agents modeled in the original position) for the
agreement of actual moral agents that is to be secured through the
employment of the model. His claim that the Rawlsian procedure aims
merely for agreement among hypothetical agents thus reflects a level
error.
Second, Rawls does not intend the parties in the original
position to represent a conception of the person. While Rawls does
intend to "specify a reasonable conception of the person as an
element" (304) of his procedure, that conception is not embodied in
the representation of the parties. Rather, as discussed above, Rawls
argues that a Kantian conception of the person views the person as
characterized by four qualities: (i) the rational; (ii) the
19
Convention