In recent years a rather different way of analysing intonation, sometimes referred to as autosegmental, has become quite widely used, especially in American work. We will look briefly at this, in a simplified account that tries to introduce some basic concepts. In the autosegmental approach, all intonational phenomena can be reduced to just two basic phonological elements: H (high tone) and L (low tone). A movement of pitch from high to low (a fall) is treated as the sequence HL. Individual stressed ("accented") syllables must all be marked as H or L, or with a combination marking a pitch movement, and with an asterisk * following the syllable. In addition, H and L tones are associated with boundaries. A major tone-unit boundary (equivalent to what we have been marking with ||) is given the symbol %, but it must also be given a H or a L tone. Let us take an utterance like 'It's time to leave', which might be pronounced

its 'time to leave (using our usual transcription)

The basic parts of the alternative transcription might look like this (the tone symbols may be placed above or below the line, aligned with the syllables they apply to):

H*      H*L%..

its time to leave

Instead of marking a falling tone on the word 'leave', the high-pitched part of the word is shown by the H and the low part by the L associated with the boundary %. There is another boundary (corresponding to the minor tone-unit boundary |) which is marked with -, and again this must be marked with either an H or an L. There must always be one of these boundaries marked before a % boundary. So, the following utterance would be transcribed like this in the system:

we .looked at the /sky | and 'saw the clouds

and in this way using autosegmental transcription:

L*           L*H-      H*   H*   L-L%

we looked at the sky and saw the clouds

How would this approach deal with complex tones spread over several syllables?

H*    L-H%

vmost of them could be transcribed most of them

Although this type of analysis has some attractions, especially in the way it fits with contemporary phonological theory, it seems unlikely that it would be more useful to learners of English than the traditional analysis.