The estimated reading time for this post is 10 minutes.
By Jeroen Bouterse
A physicist [P], a sociologist [S] and a historian [H] on demarcation in history of science.
H: “So, I understand you hadn’t heard of Shells&Pebbles yet?”
S: “I’m afraid not; what is it?”
H: “It’s a blog about the history of science.”
P: “Oh, then I have seen it! But there was hardly any science on your blog. I’ve read things about medical novels, graffiti and fern, but…
S: “Perhaps this goes to show that you have a rather restricted view about what science is.”
P: “That may be so. Call me old-fashioned, but in my view, science is an activity that distinguishes itself from other activities by searching for rational understanding and explanation of nature. So, if we are to call something scientific, it needs to conform to certain minimal requirements; for instance, that claims about nature must in the end be supported by empirical observations.”
H: “That is old-fashioned indeed. For one, what do you mean by ‘supported by observations’? The answer to the question of which claims are supported by observations has changed quite a bit over time.”
S: “Moreover, even now it is simply impossible to formulate a universal norm that states the right relation between observations and scientific theories – if those are indeed separable categories, but I will let that point pass for now. The point is that if you want to define a priori what ‘real science’ is, you are not just being anachronistic, but you are obscuring the interesting questions about the social processes that determine the demarcation of science in society.”
H: “This comes close to my second problem. In this way, you exclude every ‘wrong turn’ (at least, what you think are wrong turns) from the history of science by definition. If scholars in a given age all believe in phlogiston or in a material ether, you (P) can simply ignore this: the mere fact that these theories have been rejected means that they must have been insufficiently ‘supported’ by observations. Since at the time there was no reason to regard these theories as unscientific, history of science done from your perspective can be nothing other than what we call ‘Whiggish’: a success story, because it includes only those steps and decisions with which we already agree.”
P: “‘Whig history’ is not a label I fear. However, what you two say does not do justice to my position. There are two reasons why my criterion – the trivial criterion, I maintain, that history of science must be about something that we can recognize as science – allows for the study of theories that have later been rejected, or the study of practices and research tradition that have since perished.
First, it is perfectly okay for historians of science to investigate phenomena that are causally relevant to their proper subject matter. If Isaac Newton was a practicing alchemist, and if he drew from his alchemical work concepts, theories or even just ‘inspiration’ that turned out to be relevant to the development of his theory of universal gravitation, there is no reason to ignore this. I merely ask for conceptual clarity: the fact that a scientist like Newton spent time on alchemy or theology does not mean that alchemy or theology should be considered sciences. If Einstein liked travelling by tram, that does not make travelling by tram ‘science’; its relevance to the field of history of science must be shown by making a connection to something that is (demonstrably) scientific.
Second, it is decidedly not the case that rejected theories are qualitate qua unscientific. Also, when I said claims need to be ‘supported by’ observation, I didn’t mean to preach a naïve inductivism. Science is precisely about making falsifiable claims, testing these claims, and rejecting them if they are, well, falsified.”
S: “As far as I am concerned, this is no improvement. A hypothetical-deductivist model of science is just as much a matter of convention as an inductivist one. The interesting issue is precisely that different societies and cultures delineate authoritative knowledge in different ways. If you think you know exactly what the ‘genuinely’ legitimate ways to gain knowledge are, that is all right with me, but you need to understand that this in no way helps you gain insight in the processes through which other societies have done so. Science is an actor’s category, not an explanatory category.
H: “Our task seems to be to translate the demarcation problem from the philosophy of science to the history of science, so we should examine to what extent criteria of demarcation and arguments against demarcation are relevant to the historian. One kind of criterion that seems untranslatable is a prescriptive one: a historian as such should not judge past activities on the basis of contemporary norms – not because this would be immoral or uninformative for anyone to do, but because it is not his task as a historian.”
The historian always encounters science as it is (or was), not (necessarily) as it should be according to a certain philosophy of science. Of course, it is possible that the science he encounters is an activity that is in practice governed by certain rules or defined by certain characteristics, and it may be useful to identify those rules and characteristics independently from actor-terminology. Then again, this does not mean that the norms of current science are always helpful in identifying these rules and characteristics.
Things change. It is therefore prima facie unlikely that we will encounter a lot of activities in the past that would count as ‘science in our sense of the word’. Modern medicine may require that clinical trials for new drugs be done with blind or double-blind experiments, with the control group being given a placebo; moreover, these standards might even be rational and failure to live up to them might rightfully lead to exclusion from the scientific community (though this judgment of rationality is already outside the scope of the historian). But if they were not present in a different time or place, the same ‘failure’ cannot be a reason to deny an activity its ‘scientificity’, especially if there is an obvious continuity in all other respects between that activity and the one we now call scientific. It is evidently not the case that medicine as a whole only became scientific when it developed double-blind experiments: rather, scientific medicine modified itself so as to contain double-blind experiments as a necessary condition for inclusion.”
P: “But why the historical fuzziness? Is it really such a problem to deny the label of ‘science’ to activities that do not fit our descriptive norms? If I’m not mistaken, historians have no problem agreeing that the German Democratic Republic did not actually fit our existing norms for democratic government, and that its government only became ‘democratic’ when Germany was unified, possible cultural and economic continuities notwithstanding.
Why is it so hard to grant that, even when the historical actors we are studying considered a certain activity in the past ‘science,’ and even when that particular activity had the potential to morph into something we can now recognize as scientific, it simply does not fit our current definition of science, and that therefore this adjective must be withheld?
S: “That is interesting. Such a strict definition of science would imply that we need to exclude Newton and Boyle from the history of science.”
P: “Perhaps, my friends, you overestimate my propensity to canonize and venerate the heroes of the scientific revolution, and my readiness to swallow any absurd statement just because it contains a reference to the authority of Isaac Newton. I have said before that if Newton occupied himself with alchemy, alchemy does not therefore become scientific. I will add now, though one would think it goes without saying, that if Newton did nothing that we can recognize as science, then, yes, even Newton should be excluded from the history of science. I admit that, intuitively, I find this unlikely to apply in actuality, but intuition can fail; and I am ready to sacrifice Newton to the virtue of conceptual clarity. So, yes: if our culture decides that scientific medicine requires double-blind experiments, then it must draw the conclusion that medicine before double-blind experiments was not scientific.”
H: “Once you realize that this line of reasoning is absurd, your clear formulation will have brought us a lot closer to a meaningful account of historicized definitions, and to a breakthrough in solving the demarcation problem in the history of science. My point, after all, is almost exactly an inversion of what you just said, but with a temporal dimension in order to remove the apparent absurdity from my: I say that, even though our culture dictates that current scientific medicine requires double-blind experiments, it must not draw the conclusion that medicine before double-blind experiments was not scientific.
But let me get back to your argument first. It is courageous of you to consider sacrificing Newton – it shows that you are willing to throw out the baby if that is the only way to get rid of the bathwater. But would you also consider denying scientific status to current medicine?”
P: “Why would I do that?”
H: “Suppose that, by 2020, the field of medicine will have developed a new set of methodological standards, which significantly increase all the epistemic virtues of medicine, such as predictive accuracy, at no or negligible cost. Once this new set of norms is accepted by the scientific community – let’s for now assume it does, and among the three of us you I’m sure you would be the first to grant both that such an unambiguous improvement is possible and that it would be accepted – it will become part of the definition of scientific medicine, just like double-blind experiments. Anyone who fails to live up to it will be rightly excluded from the scientific community. And any ‘past practitioners’ (i.e. all of us now in the second decade of the 21st century) who would fail to live up to these new standards, would be rightly denied a scientific status. You have forced yourself to recognize that our medicine is not scientific. Am I not right?”
P: “Ironically, in your desire to historicize the definition of science, you have unintentionally made me the better historian, in the sense that you fail to take into account the historical situatedness of history-writing itself. I have been careful, after all, to stipulate that it is our culture whose judgments on scientificity have to inform our history-writing. I am ready to admit that a future culture may judge what we call science – or more likely, parts of what we call science – to be unscientific after all, and exclude those parts of science from their histories of science; I do not intend to be anachronistic, however, and already anticipate the perspective of this future culture.”
S: “Hear, hear; now she turns out to be a relativist after all. If you agree that definitions of science are culture-bound after all, what separates your position from a more sociologically informed demarcation of science in history?”
P: “Again, I do not fear those presumably threatening categories. If you will only allow me to stipulate what I consider to be relative to what: I have stated no more and no less than that the definition of the history of science is dependent on – ‘relative to’, if you will – the definition of science; and I have recognized that this definition may change over time. For the better, I presume.”
H: “But this is absurd! What you have said means that at any time, the history of science will be extremely short.”
P: “Maybe. But I do not exclude the possibility of a causally relevant ‘prehistory’ of science. The history of the concept of human rights as we currently understand it is relatively short, but its cultural roots are manifold and reach far back, and historians have rightly studied them. Analogously, as I said before, I do not reject the possibility of including in the historiography of science causally relevant episodes that are not in themselves scientific. But if you ask the inverse from me, namely to call these episodes scientific because they are justifiably included in the historiography of science, then I decline. You can demand that I recognize that definitions be ‘historicized’, but, each time you use that word, I cannot help but hear it as ‘obfuscated for utterly contingent and irrelevant reasons’.
Anyway, regardless of whether we call medicine-before-double-blind-experiments science, and regardless of whether we consider Newton a scientist; I do agree that they both are legitimate objects of research for the historian of science, precisely because of their undeniable causal connections to the development of science. This also accounts for applies to the continuity you previously mentioned: the stronger the causal connections between current medicine and a past state, the stronger the reasons to study this past state. There is, I take it, no disagreement between us here; if you feel the need to call this past state ‘science’, that’s unfortunate, but at this stage I do not feel the need to challenge your indulgence in fuzzy definitions any further.
H: “I am not ready to agree to disagree on this matter quite yet, for one because I, too, refuse to rest satisfied with fuzziness, and I too want to develop a way to identify with some precision the episodes relevant to my discipline – the difference is not that I intend to be more ambiguous, but that I think that a historical dimension should be taken into account, in a way similar to the way S thinks the social dimension should be taken into account. Our semantic disagreement here seems to point to a disagreement that is not merely semantic, but substantive. I have identified it before, by the way, by calling your approach ‘Whiggish’.”
P: “Again, the name-calling.”
H: “I have a problem with your claim that including past episodes in the history of science is justified by their causal relevance to current science. You have said just now that it would be conceivable to deny the label ‘scientist’ to Newton – I would point out that given the extremely present-oriented definition of science you are defending, it would in effect be nearly impossible to call Newton a scientist, and it would certainly be impossible to find someone using the procedures of Galileo Galilei in a modern department of physics. My question is: given that Galilei was not a scientist, am I to conclude that if I find a counterpart to Galilei who has done every canonical thing that Galilei has done – from defending heliocentrism to measuring the speed of balls rolling from inclined planes, to discovering Jupiter’s moons; let us, for good measure, suppose that his work was identical to that of the actual Galilei – but who has had no effect whatsoever on the history of science (because of a very effective damnatio memoriae by a provoked patron, or because of a natural disaster – whatever unlikely circumstance may have brought this about), am I to conclude that if I find such a figure, I am, by your definitions, forbidden to study him as a part of the history of science?”
To be continued…