Utterances in infant-directed speech are shorter, not slower

• Infant-directed speech is often said to be slower than adult-directed speech.
• Previous studies of speech rate have relied on global calculations over entire utterances.
• Infant-directed speech uses shorter phrases; combined with phrase-final lengthening, this results in an apparent lower speech rate.
• Speech rate differences between infant-directed and adult-directed speech disappear when phrasal position is taken into account.
• In spontaneous speech, mothers do not speak more slowly to infants.

It has become a truism in the literature on infant-directed speech (IDS) that IDS is pronounced more slowly than adult-directed speech (ADS). Using recordings of 22 Japanese mothers speaking to their infant and to an adult, we show that although IDS has an overall lower mean speech rate than ADS, this is not the result of an across-the-board slowing in which every vowel is expanded equally. Instead, the speech rate difference is entirely due to the effects of phrase-final lengthening, which disproportionally affects IDS because of its shorter utterances. These results demonstrate that taking utterance-internal prosodic characteristics into account is crucial to studies of speech rate.