replay: the trouble with thought experiments

In 1877, John Tyndall gave an address in Belfast that was emblematic of the high and confident positivism of the time. In one passage, he violates one of the canons of Victorian gentility – the Oxford variety – by aligning himself with the gloriously vulgar tradition, going back to Francis Bacon, of using Aristotle, conceived of as the father of  a lot of a priori nonsense, as an all purpose punching bag:   
“…in Aristotle, as in Goethe, it was not, I believe, misdirection, but sheer natural incapacity which lay at the root of his mistakes. As a physicist, Aristotle displayed what we should consider some of the worst attributes of a modern physical investigator: indistinctness of ideas, confusion of mind, and a confident use of language, which led to the delusive notion that he had really mastered his subject, while he had as yet failed to grasp even the elements of it. He put words in the place of things, subject in the place of object. He preached Induction without practising it, inverting the true order of inquiry by passing from the general to the particular, instead of from the particular to the general. He made of the universe a closed sphere, in the centre of which he fixed the earth, proving from general principles, to his own satisfaction and to that of the world for near 2,000 years, that no other universe was possible. His notions of motion were entirely unphysical. It was natural or unnatural, better or worse, calm or violentóno real mechanical conception regarding it lying at the bottom of his mind. He affirmed that a vacuum could not exist, and proved that if it did exist motion in it would be impossible. He determined a priori how many species of animals must exist, and shows on general principles why animals must have such and such parts. When an eminent contemporary philosopher, who is far removed from errors of this kind, remembers these abuses of the a priori method, he will be able to make allowance for the jealousy of physicists as to the acceptance of so-called a priori truths. Aristotle's errors of detail, as shown by Eucken and Lange, were grave and numerous. He affirmed that only in man we had the beating of the heart, that the left side of the body was colder than the right, that men have more teeth than women, and that there is an empty space at the back of every man's head. 
There is one essential quality in physical conceptions which was entirely wanting in those of Aristotle and his followers. I wish it could be expressed by a word untainted by its associations; it signifies a capability of being placed as a coherent picture before the mind. The Germans express the act of picturing by the word vorstellen, and the picture they call a Vorstellung. We have no word in English which comes nearer to our requirements than Imagination, and, taken with its proper limitations, the word answers very well; but, as just intimated, it is tainted by its associations, and therefore objectionable to some minds.” 

Tyndall’s groping attempt to put his chemical stained fingers around a term to distinguish a distinct, yet under-conceptualized  mental act  – and can’t one feel him almost painfully balance just on the edge of the unknown word, like Watson trying to follow one of Holmes’ points – eerily points to the need that was met ten years later, when just the thing emerged under the pen of a German physicist, Ernst Mach. The Gedanken-experiment was born. 

Ever since, it has been retrospectively accorded to other times and conceptual schemes. I have always found this a rather uncomfortable anachronism. But what I’d like to consider is how, exactly, the thought experiment is an experiment. 

We don’t kid ourselves that our objections will squelch the word. We don’t want to. The relation between the thought experiment and the experiment is like the relation between the red breasted American thrush and the English robin: they look enough alike that English settlers in the New World called the thrush a robin. Lexically, only a pedant would object to that – taxonomically, it is a disaster.  

A common defense of thought experiments, among philosophers, is that thought experiments are a common element of science. In fact,  we have read claims that in certain scientific discourses, they have an essential function. I don’t doubt it. However, the move from saying that that class of things that we call “thought experiments” play a role in science to saying that they are indeed a type of experiment is not dependent on a clear view of experiments, but on the prestige of science, which is considered to be ultimately experimental. In other words, we are eye to eye with a vicious circle. Prestige, here, underwrites this logical leap. What it tells us is two things: we are dealing, first of all, with myth; and secondly, we are dealing with myth in terms of a the archaic system of legitimation that consists in referring to authority, rather than rationality. 

Our protest against the prestige of thought experiments in philosophy stems from our sense of what experiment meant in the first place. Tyndall’s cool evaluation of Aristotle might not be textually correct re the man himself, but it is certainly correct about the spirit of Aristotelianism. The introduction of the experimental method in Europe in the seventeenth century was about one thing: the art of discovery. The point was to get outside of your head. That the world outside could be discovered was a tremendously exciting and hazardous thing. 

The mania for thought experiments cruelly inverts this moment. Reflection, instead of being forced to confront the obdurant outlines of some irrepressible piece of exteriority, contents itself with the soft and pleasing task of creating bad fictions in the image of its desires. The movement from Bacon, whose death as a ‘martyr to experimentation’ is well described by Macaulay to  the spectacle of a Chalmers, doing “consciousness science’ by means of infantile fantasies of zombies, is a painful indicator that civilization ain’t what it used to be. 

In  a conference on thought experiments that was published in the 1992 PSA: Proceedings of the Biennial Meeting of the Philosophy of Science Association, Ian Hacking, one of my favorite philosophers (who has gained this coveted status by being interested in what is going on outside of his head and studying it – a rare thing), commented on the papers presented that defended the validity of the thought experiment. He conceded the force of many of the arguments for thought experiments, but his emphasis was on the fact that he felt, in the presence of the thought experiment, unmoved. That is, he felt that the experiment was not explicative. Experiments, in Hacking’s account, have a life – thought experiments exist frozen in their pictorial essence. Referring to Thomas Kuhn’s essay on thought experiments, Hacking points to the character of good thought experiments: 
“… thought experiments are rather fixed, largely immutable. That is yet another respect that thye are like mathematical  proofs, but good proofs have proof ideas that can be used over and over in new contexts – which is not, in general the case with thought experiments. They have just one tension to expose. Of course there are false starts, and the exposition gets neater over time. And here the prescience of Kuhn’s paper comes to the fore. The reason that people wrestle with thought experiments, use them for exposition and put-down argument, is that they can reveal tensions between one vision of the world and another. They can dislodge a person from a certain way of describing the worlds. They can replace one picture by another. That is their job, their once and future job.” 
Note: Since I began this number in Victorian prose, let me end it the same way. Here is Macaulay’s great description of Bacon’s death: 
„It had occurred to him that snow might be used with advantage for the purpose of preventing animal substances from putrefying. On a very cold day, early in the spring of the year 1626, he alighted from his coach near Highgate, in order to try the experiment. He went into a cottage, bought a fowl, and with his own hands stuffed it with snow. While thus engaged he felt a sudden chill, and was soon so much indisposed that it was impossible for him to return to Gray's Inn. The Earl of Arundel, with whom he was well acquainted, had a house at Highgate. To that house Bacon was carried. The Earl was absent; but the servants who were in charge of the place showed great respect and attention to the illustrious guest. Here, after an illness of about a week, he expired early on the morning of Easter-day, 1626. His mind appears to have retained its strength and liveliness to the end. He did not forget the fowl which had caused his death. In the last letter that he ever wrote, with fingers which, as he said, could not steadily hold a pen, he did not omit to mention that the experiment of the snow had succeeded "excellently well"