This research documented a linguistic norm account of direction of com
parison asymmetry effects in relational judgments (e.g., seeing hyenas
as more similar to dogs than dogs are similar to hyenas). The asymmet
ry effect is magnified by discrepancies in prominence between subject
and referent, and has previously been explained using Tversky's (1977)
feature-matching model. Given a linguistic norm to place more promine
nt objects in the referent position, violation of this norm might redu
ce sentence clarity, which then weakens the magnitude of subsequent re
lational judgments. This research showed that clarity perceptions pred
ict the magnitude of relational judgments independently of the cogniti
ve manipulation of the features of the compared objects. The pattern o
f findings suggests that a linguistic norm interpretation may account
for variance in relational judgments independently of Tversky's (1977)
feature-matching model.