If enough data points trouble the theory, the theory may be wrong. If the theory is basic, the paradigm may have to change. Kim Stanley Robinson