On the representation of tone in Element Theory * Nancy C. Kula 1. Introduction Revised Element Theory in its current form that assumes maximally six ele- ments |A I U H L| presents interesting challenges for the representation of segmental contrasts and naturally implies a rethinking of previous assumptions. One of the basic assumptions on elements has been their ability to be indepen- dently interpretable so that each element has its own identifiable signature (Har- ris and Lindsey 1995). Thus in the representation of segments both simplex and complex expressions are permissible. Traditionally one way of extending possi- ble contrasts has been by the use of headedness (Kaye, Lowenstamm, and Verg- naud 1985) creating complex expressions where the characteristic of the headed element is enhanced in an expression. Headedness takes on an even more impor- tant and indeed slightly altered role in revised Element Theory where there are a fewer number of elements with which to create contrasts. Headedness is thus no longer viewed as the enhancement of a particular elemental characteristic within a complex expression but is rather equated to the identification of an indepen- dent signature identifying a distinct characteristic to its non-head counterpart (Nasukawa and Backlay 2005; Backlay and Nasukawa 2009). In such a scena- rio, formerly unacceptable doubly headed expressions (though see Rennison and Neubarth 1998) become fairly standard and acceptable. Similarly, there is pre- sumably nothing against an expression containing the ‘same’ element although the unmarked case is that this does not usually occur. Thus, if velar place is represented by |U | and labial place is represented by |U| these two elements may occur in the same expression in the representation of a doubly articulated labio- velar approximant. Since all elements are able to occur in both nuclear and onset positions but take on slightly different characteristics, every element has four different possible interpretations, two each in onset and nuclear position. Another way of capturing the different characteristics that have to be asso- ciated to a particular element with a reduced element set is by taking recourse to element geometries (adapted from feature geometry as in Clements 1985) in which case the different characteristics of an element follow from its position in * This paper develops from a joint presentation made with Sophie Salffner at a workshop on elements held at UCL in 2011. I thank Sophie for illuminating discussion and the audience for their input. The usual disclaimers apply.