I6 16 1 A Brief Philosophy of Laboratory and Development No one believes a theory except the theorist Everyone believes an experiment except the experimenter This often used adage although not aways accurate in detail carries a certain element of truth Here we consider certain experiments analogs models or fundamental studies of basic fluid dynamics that are intended to be relevant to one or another aspect of the ocean circulation These are physical fluid models as opposed to numerical analytical abstract or conceptual models that make use of a real fluid When the fluid in a container is subjected to some driving force the fluid moves It is observed by the experimenter It is there It is a real fluid circulation Apart from its relevance why should it not be believed The experienced experimenter is well aware of the pitfalls of his trade Aside from the accuracy of his reported observations there are many questions Were the boundary conditions well controlled Were the physical properties of the materials and their variations during the experiment known Did the methods of observation probes dyes tracers lighting influence the results Were the reported observations complete or at least representative or were they filtered massaged interpreted selectively reported etc In fact there is always an element of judgment selectivity and interpretation concerning what to observe and having observed what to record and what to report Thus despite the apparent confidence that may be exuded in publication the cautious experimenter maintains a restraint born of continual reflection on the extensive preparations necessary for an apparently simple experiment and the many opportunities for error and misinterpretation We can broadly classify experiments into four categories according to their intent under these headings The Origin of Laboratory Models and Analogues of the Ocean Circulation Alan Faller Experimentation 1 simulation 2 abstraction 3 verification and 4 extension In the first category the experimenter attempts to represent nature in miniature in so far as possible An effort is made to include all of the relevant driving mechanisms and the geometry is scaled as in nature although some distortion may be necessary Using theoretical guides such as the matching of the appropriate nondimensional ratios the intent is to learn by trial and error to what extent the ocean circulation can be reproduced as a scale model If it were possible to reproduce known features of the North Atlantic circulation for example such a model could be used to predict similarly scaled features of the ocean circulation in less accessible regions of the world The predictions would serve as a guide to further exploration and would be compared with observations as they became available The simulation mode of experimen462 Alan J Faller tation is appealing to the eye of the layman who can rather easily be convinced of its possible application and relevance The second mode of experimentation is rather like abstract art The artist draws out abstracts from some natural subject those features that he imagines to be of significance and he displays his interpretation of those features on canvas or in stone for the reaction of his peers and his public The experimental scientist conceptually isolates one or more processes that he believes to be significant in nature and he displays and tests them in the form of an operational laboratory experiment Just as it may be difficult for the artist to persuade his lay audience of the sincerity of his efforts although his fellow artists understand so also the scientist may have difficulty convincing his nonprofessional audience that his experiment is relevant to the grand scheme of things past present and future Abstract experiments may be particularly successful in systems that can be decomposed linearly without doing violence to the essential dynamics i e systems in which the abstracted phenomenon can be isolated by virtue of a lack of coupling with other processes This may be possible because of a smallness of amplitude of potentially interactive processes or because of mismatch in the temporal and spatial scales of the various processes But even in those situations where decomposition is not warranted one can say If Perhaps a planet will be found where these conditions prevail or perhaps some machinery in a chemical plant somewhere generates the conditions that I am studying or perhaps I can build upon this experiment to incorporate the interactive processes necessary for a more realistic representation of the oceanic circulation simulation In any case I will publish my abstract results for the benefit of posterity Abstraction experiments in contrast to direct simulation are more readily subject to a posteriori mathematical analysis because of their relative simplicity They may more directly lead to the advancement of theoretical aspects of the problem Either mode may be regarded as exploratory for it is likely that certain new aspects of the fluid circulations will emerge that were not anticipated and that will require rationalization Here again the abstract experiment has the advantage of its simplicity for deviations from the anticipated behavior will be more clearly recognized The simulation experiment however will generally pose a larger variety of unanticipated phenomena because of its inherently greater complexity The verification mode of laboratory experiment implies an apparatus designed to test and verify a specific analytical or numerical model A certain theoretical model predicts a steady state circulation the temporal development of a flow or perhaps an instability Apparatus is designed to match the conditions of the theory in so far as possible and as often as not the theory is modified to conform to the limitations of the experiment But in all cases under this category there is a detailed theory capable of a priori predictions and the physical conditions of the theory and of the experiment are closely matched If this matching is sufficiently precise the experiment will exactly verify the predictions From this viewpoint the situation may at first appear to be rather sterile but this is not the case at all For following the adage No one believes a theory everyone believes an experiment we find that the theorist has arrived For now everyone believes the theory and rightly so because it is confirmed by the experiment The verification mode of laboratory experiment is widely
View Full Document
Unlocking...