“One well established and generally accepted method of treating systems which are complicated beyond comprehension is to construct simple models and see whether they fit the systems in question. If they do, you will immediately become suspicious, and so will your colleagues most certainly, with the result that a blooming literature sprints up (or breaks out) dealing with the problem of how you have managed to make all your errors cancel one another. If they do not fit, the beauty of the models themselves may shine for years untainted by the squalid awkwardness of reality.”

— Kai Lang (1956), “The Thermodynamic Activity of the Male Housefly”; cited by Mogens Westergaard (pg. 289) [2]

“Cybernetics has several times proved useful when dealing with ‘borderline’ scientific problems; indeed it has sometimes been considered almost a psychiatric miracle-cure for such problems; and microbial genetics and especially bacterial genetics is certainly a borderline problem.”

— Mogens Westergaard (1957), “On the Identification of Genetic and Non-Genetic Variation in Bacteria” (pg. 280-81)

“I apologize that most of this paper has been concerned with semantic problems, since this is generally a considered a most boring subject in biology. However, I have followed a standard practice in dealing with complex problems by constructing a model. To place this model in its true perspective, the reader is referred to a recent publication by two great scientists [F. Fizz-Loony and Kaj Lang (1956)] on the use of models in biological research.”

— Mogens Westergaard (1957), “On the Identification of Genetic and Non-Genetic Variation in Bacteria” (pg. 289)