In response to Pat Gunning and Jon Medaille: I agree that "the data about human beings can be treated just as objectively as the data of physics, although a different method must be used." However, the problem as far as I see it, within the discipline as a whole, and this statement per se, is that there is rarely (if ever) paid any attention to the DATA itself. Most variables are accepted by economics as given by a neutral statistician devoid of any theoretical implications - this I feel is a grave mistake to make, and it makes a lot of econometric work and resulting theoretical work spurious and often plain confusing, as the investigator is often unaware of specifics of variables, and the manner in which the data has been a) collected b) composed and c)defined. It is true that "Your survey results will ultimately depend on where a person a million miles away puts a tick" - but not being aware of what is being ticked, or why it was, is often a problem. In continuation of the anti-orthodoxy issue raised by Pat Gunning, one might see why historians of economic thought have a chip on their shoulder in regards to the mainstream departments and their views of the history of economics. But I feel that these sentiments should not be allowed to cloud people's judgement, in reference to Jon Medaille's earlier reply to this post where he states that "Neoclassical economics are therefore, like every thing else human, value-laden;" - In fact, in answering the question, regardless of the digression about mathematics in general (which was not really asked), he should have laid this at the feet of all modern economics - as it is a trait common for all social sciences, and one which is 'under-recognised' in ours. We can treat human and physics data the same way in statistical work. But where physics will be able to define very specifically and for all cases what they mean by each variable, Economics has no such privilege. While physics should not need to worry overly much of their method for collecting that data (unless they are doing simulations) economists should always care about the method of collection and the means by which it was done - but we do not. This is where the problem lies, modern economists rarely -if ever- collect data on the ground, perform surveys or participate in raw data collection - as opposed to most of the other social sciences. This would have probably surprised many of the early writers, who for the most part relied on their own data collecting, and own definitions in order to create economic theory and statistics. Benjamin H. Mitra-Kahn