[content note: non-graphic mention of rape in second-to-last paragraph]
Thought experiments are useful.
Not all thought experiments, not in every form, not on every topic, i.e. not everything someone might call a “thought experiment” is inherently useful, valuable, or worth entertaining. A well-designed thought experiment however can increase understanding of a concept, explore new questions and perspectives, clarify otherwise murky aspects of various issues, and uncover flaws and contradictions in ideas.
Probably most important for a useful thought experiment is that something new can be learned through it. A thought experiment needs to present us with a new idea or a new perspective on an idea that lacks the biases of familiar perspectives. For example, there would be little value in re-inventing Maxwell’s demon, because this perspective on the 2nd Law of Thermodynamics is now well-explored, with much critical as well as supportive work already in existence. A thought experiment on this topic would have to introduce something genuinely novel to be worthwhile.
Thought experiments are also very susceptible to the garbage-in-garbage-out problem: a thought experiment is only as useful as the experimenter’s understanding of a topic. Premises that reflect a misunderstanding or lack of knowledge will lead to faulty conclusions, and so do badly structured experiments. And sometimes, the flawed results can become highly influential, especially when they reaffirm already held erroneous beliefs. Descartes’ rationalist (in the counter-empirical, all-knowledge-comes-from-reason sense) thought experiments stuck the world with Cartesian dualism, an incorrect worldview that infests many everyday concepts and many ideas about humans as individuals and as social actors. However, even when thought experiments have no connection to reality by design rather than through ignorance, the output is often garbage, because they are useless. Even though thought experiments are hypothetical and often quite unrealistic scenarios, the specific aspect that is to be evaluated in the experiment needs to be reflective of a real-world issue if it is to be of any use other than to entertain the experimenter.
A last essential feature of a well-designed thought experiment is clarity. Since the value of a thought experiment lies largely in providing new insights, making them inaccessible or muddled is counterproductive. This is especially so in situations where the thought experiment is already addressing a difficult issue that is e.g. very complex, or subject to strong cultural/emotional/intellectual/etc biases. A useful thought experiment reduces a problem to only the salient parts; takes principles, structures, or ideas out of their fraught or biased context and sets them into a new, neutral (because uncontroversial or else completely fictional) context; and/or creates scenarios in which empirically impossible separations of variables might be attempted. One of my favorite examples of a well-designed thought experiment is the Famous Violinist.
The Famous Violinist is a thought experiment at the core of Judith Jarvis Thomson’s essay exploring the ethics of abortion from a perspective that, in 1971, was novel to the debate over the ethics and legal status of pregnancy termination. The beauty of that particular experiment is the way in which it tosses out the cultural and religious ballast attached to reproduction, sex, and the control of female-assigned bodies that makes abortion such an intractable issue. It does so by creating a completely new context for the issue of whether or not a “right to life” supersedes the right to autonomy over one’s body: instead of “woman”, the person is “you”; instead of reproduction, the issue is a fatal illness; and sex is not anywhere in the picture at all. Similarly, it removes needlessly complicating variables from the discussion: by making the dependent dying person an adult with full human rights, the experiment shows the irrelevance of personhood to the issue. The essay also addresses many real-world scenarios and contingencies that follow both from the anti-abortionist conclusion (that right-to-life outweighs right-to-bodily-autonomy), and from the reverse. The Famous Violinist thought experiment brought a new perspective to an ancient argument, it simplified it and removed it from emotionally and culturally loaded contexts, and showed its real-world applications and relevance, making it an incredibly useful thought experiment in moral philosophy.
Incidentally, it also answers the question “Do we discuss the hypothetical intra-uterine poet, or does emotion simply close down the discussion, in either direction?”. We have discussed it already, 43 years ago. The career change from violinist to poet adds nothing to the conversation, so let’s stop beating a very dead horse, especially when doing so hurts those whose rights and bodies are being pontificated upon.
All of this is to say that when e.g. a Richard Dawkins creates a “thought experiment” that is structurally and substantively trivial but for which he purposefully uses emotionally charged examples (which are also factually inaccurate and promote harmful ideas about rape), people will criticize him and his “experiment”. And they will do so not to create taboos, or because they don’t understand its logic, or because they’re emotional. It will be because that specific experiment is worse than useless; it is so trivial it produces nothing of value, while the examples are so toxic they produce harm; it is sensible to reject that.
And doing so is not rejection of thought experiments. Rejecting useless or harmful forms of thought experiment is not rejecting thought experiments in general, not even on sensitive topics. Rejection and criticism of false premises, especially those that are already perpetuated as “common knowledge” despite their inaccuracy; of provocation for the sake of provocation, especially when the target is vulnerable to harm as a result of the provocation; and of endless rehashing of the same point over and over again is not a witch-hunt. It is not censorship, is not creating no-go zones, is not rejection of thought experiments in general. It’s the rejection of a shoddily structured and harmful attempt at edginess that contributes nothing new or valuable to public discourse on any of the topics touched upon.
– – –
 Brown, J.R. & Fehige, Y. (2011). “Thought Experiments”, The Stanford Encyclopedia of Philosophy. [web]. Retrieved from here.
 Cohen, M. (2005). Wittgenstein’s Beetle and Other Classic Thought Experiments. [book]. Oxford, UK: Blackwell Publishing Ltd. pp. 25-27
 Blanshard, B. (2014). “Rationalism”, Encyclopædia Britannica. [web]. Retrieved from here.
 Descartes, R. (1641). Meditations on the First Philosophy in which the Existence of God and the Distinction between Mind and Body are Demonstrated. Translation by Haldane, E.S. (1911).
 Thomson, J.J. (Fall 1971). “A Defense of Abortion”, originally published in Philosophy & Public Affairs, Vol. 1(1). [web]. Retrieved from here.
 Dawkins, R. (Jul 30, 2014). “Are there emotional no-go areas where logic dare not show its face?”, Richard Dawkins Foundation. [web]. Retrieved from here.
 Miller, A.F. (Jul 29, 2014). “Richard Dawkins on Date Rape vs Stranger Rape”, Ashley F Miller. [blog] Retrieved from here.