"g. , "The Establishment") . Unlike the first three "filtering" mechanisms — which are derived from analysis of market mechanisms — flak is characterized by concerted and intentional efforts to manage public information. Flak from the powerful can be either direct or indirect. The direct could include the following hypothetical scenarios: Letters or phone calls from the White House to Dan Rather or William Paley Inqueries from the FCC to major television networks requesting documents used to plan and assemble a program Messages from irate executives representing advertising agencies or corporate sponsors to media officials threatening retaliation if not granted on-air reply time. The powerful can also work on the media indirectly by: Complaints delivered en masse to their own constituencies (e. g. Male Model, stockholders, employees) about media bias, Generation of mass advertising that does the same, By funding watchdog groups or think tanks engineered to expose and attack deviations in media coverage that endanger vital elite interests. By funding political campaigns that elect politicians who will be more willing to curb any such media deviations. "

" Also other properties of O may not be properties of M. In that sense a model is more abstract than its original. Secondly, a model M will in general have properties that O does not have and that actually aid I in investigating M. What the investigator I does in modeling is formulating a problem P in terms of the original O, translate that problem into a problem P' in terms of M, obtain a solution S' of P' and translate S' into a solution candidate S of P in terms of O. If S is actually a solution of P then the modeling process was successful and can be terminated. If, however, that is not the case then the investigator I can start over with a different problem, model, means of investigation, or relation between model and original. It is worth noting that the concept of solution to a problem depends on the investigator, its purpose, intended means of investigation, time-horizon of that investigation and so forth. Model Based Reasoning What has been presented here so far essentially is due to Stachowiak, [1]. To understand how then models are used in more detail one needs to distinguish from each other the various reference modes in which the investigator puts the model with respect to the original. "