The ‘PAT Testing’ performance was intended as the culminating event of research into liveness in systhetic speech, in which a mixed context of scientific research and theatrical performance would be experienced by a live audience. The experimental materials were integrated within a performance that had to also serve as entertainment: thus a programme of performance pieces was arranged not all of which were strictly related to the prescriptive goals defined by this research. However embedded into the entertainment were three synthetic speech performances all of which were delivered by the same synthetic voice and thus any changes in the user evaluation would not be based on the voce per se, but on modifications to the framework dimensions to which the three performances were mapped. The goal was to establish the effectiveness of the PAT framework in changing the perception of ‘liveness’ in synthetic speech. The prediction was that the results would demonstrate complexity in the audience’s comparative evaluation of human-versus-synthetic and synthetic-versus-synthetic speech when challenged to define a ‘true’ and a ‘false’ voice. This would substantiate the notion of relative ‘liveness’ posited by Auslander, and thus, equally, substantiate a perception of ‘liveness’ for synthetic voices that could be subject to manipulation within the PAT framework. A broadly uniform response of computer-generated voices equating to ‘false’ and human-generated voice equating to ‘true’ forms the null hypothesis.