Clarke's Research on Humour
Alastair Clarke
Clarke conducted unparalleled levels of research during the development of his pattern recognition theory, analysing tens of thousands of instances of humour from a broad spectrum of genres and situations.
This research is split into two main categories. The first draws on material from television, radio, films, books, witticisms, cartoons, comics, stand-up, improvisation, journalism, songs, jokes, other minor genres and many situations occurring during everyday life. While countless thousands of instances were informally considered over the years, 10,000 specific instances were analysed in a single document known as The Humour Ten Thousand. This document is currently being prepared for publication and is to be made publically available on the internet in due course. Due to its substantial length (it is expected to exceed 1,500 pages when presented) the document will be published in smaller sections of 1,000 instances. Each source of humour is briefly described alongside the identification of patterns as they occur, providing a freely accessible resource for academics, students, and the general public alike.
To compensate as much as possible for personal taste, members of the public were invited to supply or suggest the material to be analysed. To qualify, any suggestions had to be currently available for purchase on the basis that the results of The Humour Ten Thousand would be more valuable if the source material could be observed by those who wished to do so. Further sources were then identified by Clarke to cover as many formats, styles and genres as possible where popular suggestion had overlooked them.
The second main batch of research concentrates on the causality of laughter during social interaction. Two main styles of observation were adopted. In the first, 1,000 instances of laughter were analysed for various factors, including the identification of pattern sources, the intensity of the response and whether the person laughing and the person at whom they were laughing were male or female. Secondly, 25 groups of three or more individuals were observed without their knowledge for the duration of 20 instances of laughter in a group dynamics test. Time elapsed between instances, proportion of group responding, intensity of response, sex of both responding group and speaker and the patterns involved were among the variables recorded during this second lot of tests.
All reasonable effort was taken to maintain sound experimental research standards. Multiple venues were selected on a number of different occasions for the research. Selectivity of instances was avoided by observing and recording the stimuli in unbroken batches of 25 instances in the first experiment (20 in the second) before a break could be taken by the researcher. To qualify as an instance both the person speaking and those responding had to be clearly visible and audible. In the second set of observations, the group dynamics test, any instance of laughter not qualifying on this basis led to the abandonment of the batch and the deletion of the data.
This research on the causality of laughter during social interaction is the first to be undertaken in the light of pattern recognition theory’s implications, revealing as it does that the causality of laughter during social interaction is the same as that in traditionally recognised forms of humour and comedy. Many interesting implications regarding other aspects of social interaction were also observed during the experiments, and the statistical analysis, presented in the third volume of The Pattern Recognition Series, Signal And Response, provides unprecedented detail regarding humour and social dynamics.
Clarke's research represents the most comprehensive analysis of humour to date and comprises a substantial educational resource for a wide range of disciplines.
