The results of a Japanese replication of the vocalic social referencing experiment are discussed in comparison with original findings from the United States. The attempt to replicate failed. It is argued that the lack of success was due to (1) Japanese mothers' attitude toward laboratory situations, an attitude resulting from relative unfamiliarity with such research and, generally, the Japanese attitude toward "the others"; and (2) greater negative effects of separating Japanese infants from their mothers. Exploratory analysis suggested that, in a laboratory situation, a correlation exists between the style of maternal affective expression and infant's mood. It was suggested that cross-cultural study of social interaction should look beyond the circumscribed context and consider the comparability of experimental factors in cultural-contextual perspective. (Author/RH)