Minimum long segments of quantity in 120 discrimination tests at the word level were compared between a binary (Test A) and tripartite choices (Test B), and with production test results, utilising the same syllable structures. In the perception test, the materials were eight kinds of bisyllabic synthetic nonsense words. These structures were used in the production test for purposes of comparison with the results from the perception test, in which three kinds of pitch and intensity variance patterns were added to create another condition. Seven Finnish and Japanese subjects participated in the two perception tests and three speakers of each language in the production test.

The results in the perception tests manifested testing methodology. The Finnish reached the minimum durational point of long vowels/consonants in less time than the Japanese, but had relatively wider conditional variations, particularly in the vowels, than the Japanese in both tests A and B, although these variations (SD) were more stable than in Japanese. The word-structural differences had more effect than the prosodic conditional differences in differentiating between short and long segments in both Finnish and Japanese in both tests A and B. The minimum long segmental durations in both perception tests were shorter than the long segments in the production test. There was a tendency for more correlation in Finnish than in Japanese between perception and production.