It has repeatedly been shown that three-term-series problems with unmarked comparatives (e.g., taller, higher) are solved more quickly than otherwise identical problems using their marked opposites (e.g., shorter, lower). Clark's principle of lexical marking accounts for these results in terms of a simpler semantic featural coding of the unmarked comparative with respect to its marked counterpart. Huttenlocher's theory of spatial imagery accounts for these same results via the subjects' mental ordering of the three terms as instructed by the problem statements. The present research demonstrated that while the lower latency of the unmarked adjective is a reliable effect, congruent ordering strategies are necessary for significant results. Subjects who order terms from unmarked to marked produce significant results for Clark's principle of lexical marking. Those who order terms in opposite direction do not. Further, it is shown that the choice of direction of ordering is itself significantly influenced by the affective value of the adjectives in context and to the individual subject.