The Swedish Research Council's definition of 'scientific misconduct': a critique

There is no consensus over the proper definition of 'scientific misconduct.' There are differences in opinion not only between countries but also between research institutions in the same country. This is unfortunate. Without a widely accepted definition it is difficult for scientists to adjust to new research milieux. This might hamper scientific innovation and make cooperation difficult. Furthermore, due to the potentially damaging consequences it is important to combat misconduct. But how frequent is it and what measures are efficient? Without an appropriate definition there are no interesting answers to these questions. In order to achieve a high degree of consensus and to foster research integrity, the international dialogue over the proper definition of 'scientific misconduct' must be on going. Yet, the scientific community should not end up with the definition suggested by the Swedish Research Council. The definition the council advocates does not satisfy the ordinary language condition. That is, the definition is not consistent with how 'scientific misconduct' is used by scientists. I will show that this is due to the fact that it refers to false results. I generalise this and argue that no adequate definition of 'scientific misconduct' makes such a reference.