From the assumptions that vowel-to-consonant formant transitions in any given VC context are (1) similar in shape and (2) scaled in proportion to the difference between the vowel target and consonant locus, we show that the implied set of scaling relationships leads to a method for estimating the loci and targets from formant data.