Adventures in Fact-Checking: Einstein Quote Edition

I spent Friday morning working on my first ever syllabus, for the discussion sections of an upper-level undergraduate research methods course. I was thinking of including a somewhat cheesy quote but I couldn’t remember who had said it: “If we knew what we were doing, we wouldn’t call it research.” A quick Google search revealed that the quote was almost always attributed to Albert Einstein. “Awesome!”, I thought. Always great to have an Einstein quote. But, a nagging voice in the back of my head pointed out that not one of those attributions was sourced.

Twenty minutes later, I’d exhausted my usual sources (Google scholar, Google books, etc.) and still no authoritative citation. An online consult with a librarian and some reference works on quotations revealed nothing, but did point me to the Einstein Archives Online. I dashed off an email and was delighted to get a response a few hours later: they couldn’t find anything in the usual places (the usual places mostly consisting of The Quotable Eisntein). They couldn’t assure me that Einstein hadn’t said it, but they had no obvious evidence that he had.

So, what to make of all this? Did Einstein say the attributed quote, but perhaps never in a formal or written setting? If not, who did? When did it first become associated with his name? I don’t know, and I’m out of time and energy to investigate. But I am pleased that I could draw on the expertise of a reference librarian, an archivist, and the entirety of the Google books and scholar databases from my desk in a handful of minutes. It’s a fun time to be a researcher.

Phenomenal Sociology

Ian Hacking is very smart. This point is not particularly contentious, I think. Few other scholars manage to bring together traditional philosophy of science, Foucault, and current social studies of science in both theoretically rich and empirically novel ways. I read some of Hacking’s later work over the past few years – e.g. The Social Construction of What?, Historical Ontology, and some of his work on the history of probability and statistics. Recently, I had the opportunity to read an early book (1983) that introduces the philosophy of natural science – Representing and Intervening. Hacking argues in that book that philosophers spend far too much time dithering on about the accuracy of representation, and not nearly enough time understanding how scientists intervene in the world. Experiments are interesting, perhaps more interesting than theories, not because experiments resolve theoretical debates but rather because experiments create new phenomena. Superconductivity might be a good example – before scientists fooled around with cooling materials down to very low temperatures and trying to pass currents through them, there were no superconductors. Once they did, there were. A new kind of thing existed – a new way for people to intervene in the world.

Hacking’s later work moves away from the natural sciences towards psychology, with a focus on classification and the creation of new types of people. Hacking looks at multiple-personality disorder and child abuser to demonstrate what he calls “the looping effect of human kinds”. These are sorts of people that did not exist before the various human sciences had diagnosed him. Which is not to say that people did not beat their children, or hear voices and behave in strange ways. Rather, these behaviors did not latch onto a way of knowing that grouped them into a certain pathology and made a kind of person out of it. And this making of a kind of person matters, because the people who end up in these kinds change who they are because of it. Hacking refers to this as “dynamic nominalism”, the idea that the names we have for the world affect the world itself (notably, Hacking confines this story to human kinds – quarks, he argues, do not so much care what we call them, although we might care. People, on the other hand, take on the labels applied to them). A more recent example might be fibromyalgia and the fibromyalgia suffered – a human kind invented fairly recently, and which is now contested (does it exist? what does that mean?). The existence of the disease – and hence the kind of person – is contested by the medical community (see, e.g. Drug Approved. Is Disease Real?). But now, organizations exist to support sufferers, drugs are approved to treat the symptoms, etc. The human kind, a new way of being, exists, and it would take another kind of fight to make it go away or replace it with new kinds.

I’m writing this post in part because I wonder if there is a connection between Hacking’s early work emphasizing experiment and intervention in the natural sciences, and his later work on human kinds in the human sciences. Are human kinds the equivalent of experimentally created phenomena?

More broadly, what are the phenomena for a discipline like sociology? I spent this morning reading a few articles from the newest issue of the American Journal of Sociology, one of the discipline’s flagship journals. Articles in AJS contain a mix of theory and observation, causal accounts and historical narrative, and so on. For example, an article on the Detroit newspaper strikes of the 1990s (Rhomberg 2010) examines the “deviant case” of the multi-union strike at the Detroit News and Detroit Free Press. The article was well-written, and interesting from both an academic and personal standpoint, as my mother worked at the Free Press during the strike. (I tried to match the account of the case with my own memory of living through the strike, and found it difficult – I was still in elementary school when the strike began, and my memories are all very local while the narrative in the article is fairly macro. If I have one major complaint about the piece, it’s that it fails to capture anything about the tensions felt by workers, or the struggles within unions about the appropriate response to management actions, etc.) The article argues for the concept of a “signal juncture”, a moment that displays how underlying trends or shifts in historical periods are progressing (in this case, the shift from the capital-labor accord to a new era of relatively weak unions that use different organizing strategies) rather than a “critical juncture”, which is actually a crucial moment for such a transformation. I wonder though, is this the kind of thing that serves as a phenomenon for sociology? If so, it seems a bit.. weak. If not, what is?

What would it mean to focus our sociological inquiries on finding new phenomena? To some extent, anthropology’s (old?) emphasis on mapping the extremes of the variation of culture has such a flavor – literally trying to determine what is humanly possible.

Anyway, I’m lacking in a strong thread to tie this ramble together, but I’ll just finish by saying that I think it would be worth our while as social scientists to read Hacking’s take on natural science, rather than relying on older notions from philosophy of science to structure our goals for a scientific sociology. Perhaps it’s not predictive theories, quantitative methods, or cultural legitimacy (or at least not just those things) that distinguish hard and soft science, but rather it’s the ability to produce new phenomena through novel methods of observation and intervention.

A Brief Borgesian Thought About Knowledge

It turns out that I absolutely love philosophers of science who like, but are somewhat critical, of science studies. I’ve recently read Zammito’s A Nice Derangement of Epistemes: Post-Positivism in the Study of Science from Quine to Latour, Ian Hacking’s Representing and Intervening, and I am now working my way through Helen Longino’s The Fate of Knowledge. All three are critics of the various excesses of science studies (especially the Strong Programme, everyone’s favorite bête noire), but all three recognize some of the theoretical and empirical insights of science studies as crucial to understanding science in action (as opposed to some sort of rational reconstruction of how science should have been or some normative program for creating an idealized science).

One of the interesting debates that Hacking and Longino both discuss is the issue of monism vs. pluralism – basically, is there one unified set of laws that coheres completely and, if known, would describe the whole world accurately, or are there multiple (potentially inconsistent) but “correct” ways of looking at the world depending on the scale you are looking at and the kind of problem you are trying to solve. I would have thought that philosophy of science would side strongly with monism, but both Hacking and Longino seem to embrace (or at least defend the possibility of) pluralism. Here’s a snippet from Hacking (1983: 219):

God did not write a Book of Nature of the sort that the old Europeans imagined. He wrote a Borgesian library, each book of which is as brief as possible, yet each book of which is inconsistent with every other. No book is redundant. For every book there is some humanly accessible bit of Nature such that that book, and no other, makes possible the comprehension, prediction and influencing of what is going on.

Another, equally nerdy way of putting this is.. there is no sufficient statistic for the entire universe smaller than the entire universe. For any given question, we may be able to find a stable relationship which can give us nice predictions and aid us in controlling specific phenomenon with relatively small amounts of information (e.g. the ideal gas law, PV=nRT), but these laws are domain specific, and so what may be sufficient information for one purpose (e.g. knowing nRT and V gives you P) will not be sufficient for some other kind of question.

What’s interesting to me about this formulation is that it foregrounds the process of problem-choice. That is, the problems we choose to investigate will determine which books in the Borgesian library we read and choose to emphasize. Knowledge of one set of books may or may not make learning another set easier. The world may seem incoherent because our shocking ability to gain some understanding in spite of the immense complexity of everything breaks down as we try to shove more and more phenomena into the same book. But how we choose our problems, how we decide what problems are important, are messy, problems that have been ignored by much of the philosophy of science – the context of justification being preferred over the context of discovery. A Borgesian view of the world foregrounds discovery (and the communal aspects of discovery at that), because we jettison the assumption of convergence on a single, book of Nature, and assume instead a collection of librarians wandering together through an endless library, discarding most books as worthless but holding onto a treasured view that give us insight into specific bits of the messy universe-library.

Incommensurable Incommensurabilities

As part of a project trying to unpack the epistemology of the market, I am reading through some overview texts on knowledge, philosophy of science, etc. Currently, I’m working through Zammito’s A Nice Derangement of Epistemes, which examines “post-positivism in the study of science from Quine to Latour”. In the chapter on Kuhn, Zammito mentions a bit about the origins of Kunh’s idea of “incommensurability“, the idea that scientific theories from different paradigms cannot be directly compared to see which is better. Here’s Zammito:

The thesis of incommensurability was enunciated simultaneously by Thomas Kuhn and Paul Feyerabend, colleagues at the time at Berkeley, in 1962. It arose out of conversations between them, but it does not follow that it ever had synonymous meaning even for the two of them. Indeed, there are grounds for the suspicion that they never shared the same idea of incommensurability. (p. 61)

I love it! Two big-shot philosophers/historians/whatnot of science come up with a similar idea, both naming it incommensurability, about the difficulty of comparing different theories from different traditions, but cannot themselves agree upon the definition of the term! In this case, don’t you think that strengthens the idea rather than weakening it?