After Construction/Between Loops and Kinds: Alexander's The Mantra of Efficiency

My favorite thing about this blog is that it sticks things deeply in my craw, and I cannot pull them out, so now my craw is full. Today, I'd like to return to a discussion we were having several months ago about ontology (here, here, here, and here). It's never let me go. I'd like to return to it by considering Jennifer Karns Alexander's 2008 book, The Mantra of Efficiency: From Waterwheel to Social Control

The Mantra of Efficiency won the Society for the History of Technology's Edelstein Prize, for "an outstanding scholarly book in the history of technology published during the preceding three years." Like all Edelstein-winning books, Mantra was the subject of a special conference session, this one at SHOT's 2010 meeting in Takoma, WA, which included commentators like Bill Storey, Wiebe Bijker, and Tom Misa. At one point, the discussion took an unexpected turn. Misa said that Alexander's account of efficiency pointed to a post-constructionist history of sci/tech. This excited Misa (and me!). Bijker pushed back, standing up for the continuing relevance of constructionism, which he did again this year when he accepted the DaVinci (lifetime achievement) medal at SHOT's 2012 meeting in Copenhagen. Alexander, I think, did not want to wade into these dangerous waters and, as I recall, was hesitant to say that her book meant to go there. But I agree with Misa: Mantra does go there—or at least it points one way.

First off, The Mantra of Efficiency is a good book, and everyone should read it. It has crossover interest for historians of science and technology and shows a way forward in bridging that divide. If I bumble things with my butter fingers in what follows, you should not blame it on the book. Second, I'm going to give a brief account of Alexander's overall argument, but I am centrally interested in her theoretical framework, which is contained in her intro and conclusion. Moreover, although I'm interested in Misa's point about post-constructionism, I'll spend most of my time at the end using Alexander's account of efficiency to question Ian Hacking's distinction between "natural kinds" and "human kinds," which we've already tackled here in those posts on ontology I listed above.

Alexander covers the emergence and transformation of efficiency across considerable time (1750 or so to the present) and space (the US, UK, and France to the globe) by way of several case studies. (The book's form is not unlike Daston and Galison's Objectivity in that way.) Alexander is fascinated by efficiency's dual nature: there are static and dynamic visions of the notion; it is simultaneously descriptive and normative. This quotation captures the idea's double image, "At times [efficiency] emphasizes stability and conservation; at other times it is embedded in a rhetoric of transformation. It appears to be merely a technique of quantification, yet it also appears as the goal towards which quantification is employed. It can be both a model of a well-controlled process and a tool to help achieve that control." (4)

These thoughts lead Alexander to look near and far for the workings of efficiency. She shows how Darwin and Marshall used the notion to discuss, respectively, evolution and markets, and she examines how Taylorists and others used efficiency to remake work. When she talks about the latter cases and describes her own negative experiences with efficiency, Alexander almost sounds like a member of the Frankfurt School criticizing "instrumental reason." Almost but not quite—she's more timid in her moral judgment. (This issue reminds me of Adam Curtis's [highly problematic] history of how we've used machine metaphors to describe nature.)

Chuck Darwin: "natural selection [is] . . . efficient in so high a degree."
In her search for a way to describe efficiency's history, Alexander tries to clear away some previous ways of dealing with similar notions, which she thinks are problematic: "The temptation, when confronted with such a broad array of uses [of the word], is to treat efficiency in one of three ways: as the product of a specific localized context, as the metaphorical application of an engineering term, or as one among many instances of quantification." (4) Anyone attuned to the historiography of sci/tech will see that Alexander is taking swipes at some heavies—and questioning some constant themes (theories? concepts?): local knowledge, metaphor, and quantification/numbers.

This post is already going to be long, so I need to be as brief as possible here: I think that Alexander's book forms part of a critique of the dogmas of science and technology studies, which I hope to outline here and in some white papers over the coming weeks and months.

There is good reason to doubt the STS dogma that all knowledge is local, just as it is irresponsible to teach uncritically works, like Donna Haraway's essay "Situated Knowledges: The Science Question in Feminism and the Privilege of Partial Perspective," without seriously interrogating their assumptions in front of students.  For instance, you could say that Srinivasa Ramanujan would not have begun working through mathematics if it wasn't for local contexts. These local contexts would include his private life as well as the colonialism that brought books, like S. L. Loney's works on trigonometry, to his door. But this isn't saying much of anything at all. It doesn't help us determine whether his math was any good, nor does it help us understand why he was able to rediscover mathematical principles that weren't in such books. This brings us back to the old criticism, which holds that STS too often puts all of the causality on the side of the "social" rather than also copping to the force of ideas themselves. And this is Alexander's point, "Subsuming efficiency in context suggests that it is a shell, ready and waiting to take on the values and objectives of whoever uses it but with little content or character of its own."

We should also follow Alexander in questioning the academic love affair with "metaphor." (Trace the "metaphor" and "rhetoric" entries in the index of Zammito's A Nice Derangement for a good time.)  The metaphor of "metaphor" in cultural history is a good example of finding a partial truth and running with it. Are some statements and image in science metaphorical? You bet. But just as Ludwig Wittgenstein could list many different uses and goals for language (aphorism 23 here), statements and images in science play many different roles. (Sergio Sismondo has tried to create a varied account of scientific propositions and images, in his case maps, here.) For instance, if I decide to stock my farm with Hereford cows because they use calories more efficiently than Holsteins, am I somehow enacting a metaphor? That makes no sense to me.

Alexander's reason for saying that efficiency is more than metaphor sounds a lot like her point about context, "Speaking of metaphor suggests that efficiency has no substantial content of its own when used outside of the physical, material domain of engines and energy. Treating efficiency as a metaphor sidesteps the thorny question of how a technical concept might effectively be applied in nontechnical circumstances and undermines the possibility of recognizing serious attempts to extend a powerful and generalizable concept beyond the boundaries of physics or engineering."

Fred Taylor: All-Around Good Guy, Graduate of Stevens Institute of Technology. Woot!

Alexander does not say a great deal about why quantification is a bad way to view efficiency; neither will I. But the important points are captured in the bit I just quoted: quantification is an essential part of the technical definition of efficiency, but the term has become more than technical and is used qualitatively in all kinds of circumstances, including everyday life.

While Alexander deals with the pre-modern history of efficiency, including cameos by Aristotle and Aquinas, I think she wants to claim that efficiency rose above local contexts and spread so far in the world because it was/is so central to modernity and modernization. (Do you smell Adorno and Horkheimer skulking close?) This account itself can feel a bit dated, since many people have moved beyond interest in modernization theory and have become massively bored with postmodernist ramblings about modernity. But I think that some very interesting questions lurk here, and everything Alexander says could be brought inline with currently fashionable topics and movements, like the NEW HISTORY OF CAPITALISM (for our account, see here and here). Along these lines, I would have liked to have heard whether Alexander thinks, like Frederic Jameson, that there is A Singular Modernity (read "capitalism"), or if she is more of a multiple modernities-type thinker. This distinction has important implications for how we think about and describe efficiency.

Just for the record: the #1 image in a Google Image search for "efficiency"
Now to turn finally to Ian Hacking: in our previous discussions about ontology, we focused on Hacking's distinction between natural and human kinds. Hacking says that there is something special about human kinds—such as ideas about sexuality, mental health, race, etc.—in that they "loop" back and end up affecting human behavior. He believes this means that there is something particularly historical about human kinds that doesn't apply to natural kinds (electrons, species, skeletal structure, you name it). We historians at American Science were all in agreement that natural kinds are more historical than Hacking claims, though we did not sort out what this means before we ran out of steam.

I want to assert tentatively that the notion of efficiency (as Alexander explains it) calls into question Hacking's human vs natural distinction: efficiency is irretrievably and simultaneously natural and human (in Hacking's sense; does this mean something like descriptive and normative?). No one would doubt that efficiency is a human kind or that it doesn't "loop." The idea has completely remade human life and work over the last 150 years. Yet, it's a also a term we use to describe natural processes, including, for instance, digestion. Some people might want to say that such descriptions are "just metaphor." I don't think we want to go that route. Efficiency has a great deal to do with how species find, assimilate, and use calories, and (we think) it plays an important role in evolution. If that's not "real," I don't know what is.

Now, I could be getting into problems all throughout this post, but I am most aware of possible confusion right here. Lukas wrote about debates over whether dinosaurs had feathers. Perhaps we want to say that dinosaur feathers are a "natural kind" in a way that the efficiency of a biological process is not. Perhaps we want to insist that natural kinds pick out things, not processes. I might allow that, but I don't think that this assertion would deny that efficiency poses problems for Hacking's natural vs human kinds. It would just mean that we need to come up with a third kind altogether. I lack the vocabulary and theory to sort this matter conclusively. If someone wants to tell me why processes are or aren't natural kinds or desires to make a case for Hacking, please do. 

Regardless of how this theoretical discussion shakes out, The Mantra of Efficiency helps us question received wisdom in STS, including the history of sci/tech. For that as well as its lovely case studies, we all need it.


Lots of stuff here, Lee – thanks for it. Let me be try to be an orthodox Hack(-ing) again.

You're right to sense that some might want to limit ontology (and thus the designation of kinds) in a way that keeps out efficiency, though I don't think that's because efficiency is a "process." I take process philosophy to be an argument that processes (not e.g. static objects) are the only suitable answer to the (narrow) ontological question "What is there?" but that efficiency isn't a process in the sense they mean.

I'm not going to look it up again, but I think it would be safe to offer the following as either an interpretation or a friendly amendment to Hacking's account: all human kinds are natural kinds, but not all natural kinds are human kinds. Thus, if we accept "efficiency" as a kind (I don't think we should – I think it's more like a relational quality – but anyway..), it always-already disturbs the boundary between the two.

Lastly, on metaphor: "Speaking of metaphor suggests that efficiency has no substantial content of its own" seems totally wrong, doesn't it? It seems to imply we can use metaphors wherever and whenever we want ("no substantial content of its own"), which runs counter to at least my experience. I defer to Lukas on the relationship between models and metaphors, but this seems to be the weakest link here..

Hi, Hank.

You might be right about process philosophy. Learning more about it is on my short term to-do list (outside of Deleuze, whose work I know well). Alfred North Whitehead's _Process and Reality_ is an absolutely miserable book, but I've heard the place to go is his _Science and the Modern World_.

I'm going to leave aside your point about "relational qualities," though you might be onto something.

It seems to me that you have to be at least partly right about metaphor. Wittgenstein (and Cavell) spent a lot of time talking about how we extend words and metaphors to new cases. The degree to which people see the metaphor as fitting the new case is the degree to which the metaphor moves us, and, of course, metaphors can have quite powerful force; they do have their own content.

Yet, I still think it is wrong to say that efficiency is merely metaphor; it is a lot of other things, too, even if we just consider it in speech. This is Alexander's point: that efficiency should not be reduced to metaphor.

It should be reduced to mantra!

Note: Only a member of this blog may post a comment.

back to top