PURPOSE: To constitute the title device so that a reproducing sequence of a voice data group can be set simply to disorder and especially, it can be allowed to conform to response training of a conversation by determining the sequence of the voice data group to be reproduced, based on a random number which is generated by a random number generating means.

CONSTITUTION: When the title device is switched to 'random access mode' by operating a prescribed push-button of an input part 2, a random number generating means 13 generates a random number, based on a signal from a control circuit 12, and it is supplied to the control circuit 12. As a result, the control part 12 gives a sentence number corresponding to the random number which is received from the random number generating means 13, to a voice synthesizing circuit 14, and the voice synthesizing circuit 14 reads out a voice data group corresponding to the given sentence number from a solid-state storage circuit 9 of an IC card, and also, synthesizes an analog sound signal Sa, based on its voice data group and applies it to an amplifying circuit 15, and in accordance therewith, a reproducing sound is outputted from an earphone 7. In such a way, this device can set a reproducing sequence of a stored sentence to disorder and especially, can be allowed to conform to conversion response training of language study.