I was ranting during a conversation/debate with a family member’s friend when a thought occurred to me (apparently even I can’t stand listening to me). Can we dismiss pseudoscience and fringe theories and ideas as such, yet still gain from them? The reason I had this thought, apart from my apparent distaste at the prospect of paying attention to what I have to say, was a thought triggered while trying to explain the problems with a paper on “specified complexity” by Dembski (a creationist/ID proponent). I realized that although both the notion and his theory fail, some of the ideas relating to measures of complexity were not without merit. In particular, I was reminded of how much of modern statistics developed in the context of eugenics, for eugenics, and by eugenicists. Sad though this be, if these mathematical developments (divorced from the eugenics context) are clearly regarded as very useful across disciplines. Likewise, however unscientific creationist/ID arguments regarding complexity may be, it is certainly possible that mathematical components can be useful once suitably removed from the original context. It then occurred to me that perhaps the only unifying factor among the many examples of wholly or partially abandoned ideas/approaches in the sciences that in some way retain a degree of utility is being previously generally accepted or widely supported. After all, plenty of widely supported ideas/theories ( e.g., phlogiston) have left us with nothing, nor is there a unified way in which what remains of discarded notions are used. The reason outdated integration methods are still widely used is pedagogical (we can’t teach measure theory in intro calculus), while outdated (classical) physics is used because it is approximately correct in many application. Meanwhile, wholly or partially abandoned approaches to strong AI along with theories/models in biology and linguistics are widely used in contexts never intended (from pattern recognition to combinatorial optimization). Other examples are easily found, but my question is that if so horrific and unscientific a theory as eugenics can be discarded yet leave us with much of modern statistics, might not theories that are not or were not ever more than pseudoscience or fringe theories contain something of use? If so, could we separate the wheat from the chaff?