This thread grows out of an earlier discussion started by Huon Wardle, Whatever Happened to 'Social Science'? In that thread several of us who frequently disagree in all sorts of ways converged on enthusiasm for Michael Agar's new book, The Lively Science: Remodeling Human Social Research. Here we will read and discuss that book. In the spirit of the book itself, everyone is welcome to participate, but especially those who are just beginning to think about anthropology, the history of social thought, and social science. To kick things off, I offer the following short review.

-------------

The preface to Michael Agar's The Lively Science: Remodeling Human Social Research begins with the following paragraphs.

"What’s a nice reader like you doing in a book like this? I’m hoping that you’re here because you’re curious about a way to do “behavioral science” or “social science” that will help you figure out a problem you’d like to solve, or maybe you just wonder what those words mean because you’re a curious type. Maybe you’re a student, new or returning, embarking on a course with those names attached to it, or maybe a course in one of the many other areas that make use of them. The point is, I’m writing for readers who are fresh to the concepts, not for colleagues. 

"This book has a simple premise to get you started. The premise is, research on humans in their social world by other humans is not a traditional science like the one created by Galileo and Newton . It’s not that the creators were wrong. Far from it. The ones who were wrong were the historical figures who tried to imitate the way the creators worked, neglecting the fact that learning how people make it through the day is different from dropping balls from the Leaning Tower of Pisa or getting hit on the head by falling apples. Galileo didn’t have to communicate with the balls. Besides, he didn’t have to worry that the balls might look down 185 feet and refuse to jump and throw him over the parapet instead."

Two points are vital here. One concerns how we read, the other what the book is talking about. When Agar writes, "The point is, I am writing for readers who are fresh to the concepts, not for colleagues," he is asking those of us who are or hope to be colleagues to turn off what we think we already know and approach those concepts he mentions with a fresh, innocent gaze that makes no assumption that we know what he is talking about. He asks us to be readers who act like ethnographers, putting aside what we think to attend carefully to the people whose lives we share, looking for evidence of ideas that may be radically different from those we bring to the field. This is no small request, since, as indicated in the second paragraph, the topic sounds awfully familiar.

Anthropology, or at least the anthropology called social or cultural anthropology, is split down the middle. On one side are the "scientists" who see their goal as contributing to the kind of science conceived by Galileo and Newton, a science that discovers mathematical laws that work apply everywhere, regardless of what the entities they describe might be thinking or feeling. On the other side are the "humanists," for whom the essence of humanity lies in what humans think and feel and insist that thoughts and feelings cannot be understood scientifically. They can only be interpreted, thickly described in ways that make human stories plausible. Our usual reaction to this divide is to pick one side or the other and become fierce advocates for our choice. Agar asks us to question the ways in which we conceive of scientific and humanistic understanding, to challenge the divide and consider an alternative view in which science and humanity are combined. 

Stated so baldly, the thesis of the book sounds like a familiar sort of Hegelian dialectic: Thesis=science. Antithesis=humanity. Synthesis=A reconciliation that overcomes the initial contradiction. But there is much more to The Lively Science than is captured in this formula. Agar leads us on a picaresque journey through the thickets of modern social theory. He leads us away from the heavyweights usually featured in brief histories of social theory: Mars, Weber, and Durkheim. Instead he directs our attention to German idealists with names like Dilthey, rarely mentioned except in footnotes, and invites us to consider what they were on about when distinguishing naturwissenschaft, literally "natural science," from geisteswissenschaft, "spiritual sciences," a.k.a., humanities, but insisting that both are wissenschaft, i.e. science. Via this journey, Agar leads us to consider a broader view of science, in which Galileo and Newton represent only one variety, and the humanities are human sciences, retaining scientific rigor. 

As illustrated by the two paragraphs quoted above, however, this is not a pretentious, ponderous book. Agar is a witty and genial writer well-aware that his mission is not to speak to those already his colleagues, already set in our intellectual ways, but to newcomers, still young and open enough in mind and spirit and still rebellious enough to want to challenge their teachers, to consider thoughts not heard, or heard not nearly enough, in classrooms. If you see yourself in this description, please join us in reading and discussing Michael Agar's The Lively Science.

Views: 1434

Replies are closed for this discussion.

Replies to This Discussion

Michael,

Thanks so much for joining us. As Lee may have told you, at this point, we are planning to formally open the forum on October 8, to give those who don't yet have a copy of The Lively Science a chance to get one.

Note to others: the ebook is readily available at, for example, http://www.amazon.com/The-Lively-Science-Remodeling-Research/dp/162...

A PDF describing the book can be found at http://www.ethknoworks.com/files/The_Lively_Science_review_by_Floor...

To learn more about Michael Agar point your browser to http://www.ethknoworks.com/index.htm

Now, while I have everyone's attention, I would like to float an idea. As originally envisioned, this forum would be devoted to reading and discussing The Lively Science chapter by chapter. Since, however, Michael has just provided us with a beautifully clear exposition of the central argument, a different approach may be possible. Following Huon's suggestion that Kristian bring some of his ethnographic data to this thread, how would it be if others did the same?

Thus, for example, I will be talking this weekend in Taipei about "Social Network Analysis with Chinese Characteristics." The topic is not a finished piece of research, and the talk will be some very early stage ruminations. Basically as follows.

First, an ethnographic observation. I am met at an airport in Shanghai by a young woman waving a sign with my name on it. Nothing odd there. Happens all the time to people invited to conferences or business meetings. My curiosity is aroused, however, when I ask the young woman if she is a student of the professor who organized the talks I am giving. I discover that she is, in fact, the currently unemployed younger sister of a classmate of the assistant professor who has been put in charge of handling administrative details. Note, too, the talk I am giving will be in Jishou, in Western Hunan, another thousand miles from Shanghai. The young woman has been asked by her sister to help me negotiate the two-hour subway trip from the airport at which I landed to a second airport from which I will fly to Jishou. It strikes me that what just happened is something I wouldn't expect in the U.S.A. or Japan. Sure, if a friend asked me, I might go to an airport to meet them. If they were flying to Washington and might ask my daughter who lives outside the city close to both Reagan and Dulles airports if she would be willing to meet them. She might do it if she were free, but she usually isn't, so she would probably say, "Have them call Uber. You can give them my cell phone number in case they have a problem." How, I wonder, was it possible for the assistant professor to contact a classmate who then turned to her sister who met me?

I am wondering how to construct falsifiable empirical propositions to test my intuition that Chinese social networks involve chains of obligations that can stretch out three, four, or more links from the person who needs a favor, while in the U.S.A. and Japan it is hard to imagine a chain of favors more than one or two links long....

I could run on about this. My though is that by looking at a few cases we could learn more than by simply reading and discussing the book.

Thoughts?

and, behold, it was very good...

I have got as far as reading the introduction and have enjoyed it very much. The current 'crisis of replicability' in what Agar calls BSS research and the challenge posed in the Lively Science seem to be very much of a piece and to the point -- in the other thread we were looking amongst other things at the unfolding crisis of people travelling across European borders and how anthropologists might explain and understand it. This can potentially be an important case study to think through. Also we looked at another piece by Michael Agar around complexity theory. I suppose I would like to find out more about whether or how we can achieve a scale (or set of nested scales) where 'big' theoretical problems like global complexity can be seen to interact meaningfully with 'small' scale problems like the situation in a Dresden refugee centre that Kristian has been annotating. I say this in part because I find some of the current blanket concepts--'globalisation', 'neoliberalism', 'financialisation' frankly unsatisfying in as much as they obscure as much as they explain.

Complexity is a fascinating topic. The word "complexity" is, however, ambiguous. The technical usage is not, I suspect, what most of us have in mind when we describe problems as big or little in terms of geographical scale. For future reference, I reproduce here, something I wrote on Savage Minds.

---------------

Phase transitions are not always unpredictable or irreversible. H20 is a classic example here, shifting back and forth between ice, liquid, and gas, depending on pressure and temperature. The mechanisms involved are predictable.

Complexity is not random. Melanie Mitchell makes a big point of this in her introductory course at Complexity Explorer [http://www.complexityexplorer.org]. The equations that describe complexity are well-defined and contain no random variables. The problem is that solutions multiply too rapidly to keep up with them. This is related to another characteristic, sensitivity to initial conditions. One set of conditions may result in a single, well-defined solution or a relatively simple oscillation between two or three solutions. Another set of conditions, apparently only trivially different, results in a very large or infinite set of solutions. ["Solutions" is used here in the mathematical sense in which, for example y=x^2 has two solutions, x and - x. x times x equals x^2 and so does -x times -x.]

The relationship of conscious actors to complex systems is difficult to model. In Complex Adaptive Systems: An Introduction to Computational Models of Social Life, John Miller and Scott Page remark that in most agent-based simulations the agents are either too stupid — always using the same heuristics — or too smart — always performing game-theoretic calculations — to be realistic representations of human actors. It is easy to say humans learn, humans forget, humans rarely employ game theory. But the mathematical and computational issues raised by trying to incorporate more human characteristics in agents are very difficult. Modeling the human remains an on-going enterprise.

In thinking about new ideas like phase transitions and complex systems, I always keep in mind a diagram described by Gerald Weinstein in An Introduction to General Systems Thinking. The diagram is a simple box divided into three layers, two very small, the other very large. One small layer is the zone of classical mechanics, where principles governing simple machines are sufficient to explain what is happening. The other small layer is the zone of thermodynamics and quantum mechanics, where the global properties of systems that involve large numbers of randomly interacting parts are well understood. The very large zone is the one where neither simple mechanisms nor statistical inference are sufficient to explain what is happening. This is the zone in which the best we can do — to date— is to construct narratives, telling plausible stories.

The “to date” is important.The simple mechanical and statistical layers that Weinstein describes writing between 1961 and 1975 are no longer the only games in town when it comes to mathematical or computational analysis. As indicated above, there is still a lot of work to be done, but new advances continue to appear. All of which leads me to conclude that the late 1960s world in which I went to graduate school, when the choice was physics envy vs literary subtlety is no longer a good model for framing anthropological debates.

Thanks for the clarification John. The point is that 'big' and 'small' scale problems are here set up from a human perspective:--human beings find themselves in a world which is 'big' and in which they are 'small'. However naive this way of scaling things is from a certain kind of scientific stance it is fundamental to how human beings approach their relationship with the world; so the question is then how do we find ways of scaling the problems of complexity and transition using rigorous metaphors etc that are somewhere between being 'too stupid' and 'too smart'--that make some sense when we are trying to put a scale on human problems. 

The relationship of conscious actors to complex systems is difficult to model. In Complex Adaptive Systems: An Introduction to Computational Models of Social Life, John Miller and Scott Page remark that in most agent-based simulations the agents are either too stupid — always using the same heuristics — or too smart — always performing game-theoretic calculations — to be realistic representations of human actors. It is easy to say humans learn, humans forget, humans rarely employ game theory. But the mathematical and computational issues raised by trying to incorporate more human characteristics in agents are very difficult. Modeling the human remains an on-going enterprise.

As of being a social science, why not concentrate primarily on modelling the social variables in human systems? Take for instance traffic; even though full of everyday slips, it's a good example of a relatively stable human system which emerges pretty smoothly  by itself from the institutionalization of specific laws, regulations and conventions, ultimately kept in place by a collective experience of the specific sanctions entailed by transgressions. You don't really need to scale it down to individual agency to explain it. It's enough to know that, for all its dasein contradictions, ambivalence, inconsistencies etc., human behavior will ultimately converge on motivations of socially conscientiousness (the epitome research question is why is it so?). So in terms of modelling human systems in anthropology, we'd primarily be interested in social power; that is, the character of empirically observable social rule systems, and in particular in how they are enforced, experienced and acted upon at the ethnographic level. We will naturally find, as scores of anthropologists have before us, that dissent, resistance, subversion etc. is endemic to human life everywhere. That being so, writ large it nevertheless remains socially exceptional. Our interest on this end, then, would be if and how dissent feeds back to alter the game rules. To return to the traffic, in modern nation state societies, rule breaching mainly seems to serve the provision of political justification for ever tighter regulations. I'd also add that cases of dissent that don't feed much back into institutionalized systems from which they depart, as in rule breaching outside of sanction's reach (i.e. speed driving in Siberia), it's probably not that interesting from a social scientific point of view.

KRISTIAN: As of being a social science, why not concentrate primarily on modelling the social variables in human systems? Take for instance traffic; even though full of everyday slips, it's a good example of a relatively stable human system which emerges pretty smoothly  by itself from the institutionalization of specific laws, regulations and conventions,

You have obviously never been in a car approaching a traffic junction in Beijing <S>. The degree to which social relations create/develop toward stable equilibria (institutions) can only be worked out on a case by case basis and if they don't work that way it doesnt mean they are not 'social'. I entirely agree that the build up, and sedimentation, of rules for doing things is one important part of what anthropologists observe systematise and shouldnt be overlooked. I don't think it rules out a theoretical encounter with 'chaos' or 'complexity', though.

I've found it clearer to talk in terms of "nonlinear dynamic systems," a kind of formal representation that offers an alternative to the kind of linear, static, oversimplified causal formalisms traditional in human social science. Interactions and path dependence are among key concepts. When I did ABMs in my drug days I presented it to socsci audiences as a thought experiment laboratory for a structure/agency argument, but most definitely not a "flight simulator," i.e. a realistic model of a world. That exercise helped show that you could plausibly explain an illegal drug epidemic on a business/consumer evaluation model as well as you could on an infectious disease model. One way I enjoy thinking about the concept in a reflexive way is how we used to say "emergence" all the time in grad school to explain why we couldn't write a research proposal in advance of what we would learn after the research started. Legends of Kroeber's advice to just buy a notebook and a pencil as preparation were told by way of reassurance. This is emergence as phase transition, i.e. who knew in advance that would happen, rather than emergence as orderly production of pattern at a system level that could not have been predicted from characteristics of its elements. Another way to look at this topic is to go back to the anthro pioneer, Steve Lansing's work on Balinese water temples and the overview he wrote on the complexity concept for the Annual Review of Anthropology when he was permanent faculty at the Santa Fe Institute. It's a large topic, fragments of which pervade The Lively Science, I hope in appropriate ways, but I'm sure not always, and maybe not even usually. SFI is dominated by physics. It came out of Los Alamos. NECSI, the New England Complex Systems Institute, is more connected with social science type issues. CRESS, Centre for Research in Social Simulation, at Surrey University, has been a center for this kind of work for a long time. Nigel Gilbert, the director, edited until recently the e-journal JASSS, Journal of Artificial Societies and Social Simulation, worth a look if you'd like to see the different kinds of research applications. Lots to talk about here, especially given Waldrop's line in his history of the Santa Fe Institute, a book simply called "Complexity," that it represents a "qualitative holistic math." That book, though dated, is a good readable intellectual history of how the U.S. version developed.

In any case, we will get no points for reinventing the wheel. Models of traffic flows are standard items in engineering, architecture and urban planning curricula. The question is what can anthropologists, as ethnographers or as generalists with a grab bag of comparative knowledge contribute to improving what people have already done.

Lee, for example, might be interested in thinking about why, if traffic is an example of behavior that is, in most cases, in line with prescribed norms, the car chase in which all of the normal rules against breaking speed limits, reckless driving, etc., has become a stock scene in action movies and projected onto a dystopian future in the Mad Max franchise. Why are drag racing, demolition derby, NASCAR and Formula 1 hugely popular, at least with certain segments of the population?

OK; I now have laid hands on a copy of the book and as I read I see that the relevance of 'complexity' and 'chaos' is laid out clearly with regard to making social observations. So, I am off to read the rest of the book and back later.

Michael,

Thanks for all the pointers. Learned a lot here. I've known about the Sante Fe Institute for years but was unaware of the other places you mention.

Nonlinear dynamic systems, interaction, and path dependence are definitely the way to go. Just want to make another quick connection here for the rest of us: "complexity" in the mathematical sense derives directly from the study of physical nonlinear dynamic systems, e.g. weather. For an example of how such a system can be visualized, see recent weather reports on the likely paths of Hurricane Joaquin, where multiple models are being constantly updated to account for new twists and turns.



Michael Agar said:

I've found it clearer to talk in terms of "nonlinear dynamic systems," a kind of formal representation that offers an alternative to the kind of linear, static, oversimplified causal formalisms traditional in human social science. Interactions and path dependence are among key concepts. When I did ABMs in my drug days I presented it to socsci audiences as a thought experiment laboratory for a structure/agency argument, but most definitely not a "flight simulator," i.e. a realistic model of a world. That exercise helped show that you could plausibly explain an illegal drug epidemic on a business/consumer evaluation model as well as you could on an infectious disease model. One way I enjoy thinking about the concept in a reflexive way is how we used to say "emergence" all the time in grad school to explain why we couldn't write a research proposal in advance of what we would learn after the research started. Legends of Kroeber's advice to just buy a notebook and a pencil as preparation were told by way of reassurance. This is emergence as phase transition, i.e. who knew in advance that would happen, rather than emergence as orderly production of pattern at a system level that could not have been predicted from characteristics of its elements. Another way to look at this topic is to go back to the anthro pioneer, Steve Lansing's work on Balinese water temples and the overview he wrote on the complexity concept for the Annual Review of Anthropology when he was permanent faculty at the Santa Fe Institute. It's a large topic, fragments of which pervade The Lively Science, I hope in appropriate ways, but I'm sure not always, and maybe not even usually. SFI is dominated by physics. It came out of Los Alamos. NECSI, the New England Complex Systems Institute, is more connected with social science type issues. CRESS, Centre for Research in Social Simulation, at Surrey University, has been a center for this kind of work for a long time. Nigel Gilbert, the director, edited until recently the e-journal JASSS, Journal of Artificial Societies and Social Simulation, worth a look if you'd like to see the different kinds of research applications. Lots to talk about here, especially given Waldrop's line in his history of the Santa Fe Institute, a book simply called "Complexity," that it represents a "qualitative holistic math." That book, though dated, is a good readable intellectual history of how the U.S. version developed.

 

    With our “ethnographic turn” in preparation for diving into Michael’s book, I thought to return to a piece, a “mini-case study” which I posted in the old OAC forum, “From the Center for Peripheral Studies” on March 1, 2014.  It seemed relevant to Kristian’s description of traffic as a human system “which emerges pretty smoothly  by itself from the institutionalization of specific laws, regulations and conventions” and to Huon’s rejoinder about street corners in Beijing.  So, without further ado:

 

Mini-Case Study:  The California Roll

    The California roll is one of southern California’s contributions to international cuisine (taking its place alongside other classics such as the teriyaki taco and the date shake).  Southern California is also noteworthy for having probably the most four-way stops in the world, that is, an intersection with a “Stop” sign on every corner.  Every driver is required to stop before proceeding through the intersection.  Because there are so many of the things, and because many California drivers encounter literally dozens of them every day, elaborate practices have grown up to negotiate them with the least trouble.  Principal among these is the “California roll.”  As you slow down approaching the intersection, you note that no other drivers are there; yours is the closest vehicle to the corner.  Therefore, you execute a California roll: You slow almost to a stop but then r-o-l-l through it.  This eases traffic congestion that would be caused if you came to a dead stop, thus impeding your own progress and that of another driver approaching the intersection, who could have executed his own California roll moments after yours, but now had to stop. 

    Things get more complicated if there are two, three, or four cars at or very close to the intersection.  The state of California Driver’s Manual makes it seem so simple and clear-cut.  It’s not.  The Manual states that if two cars stop at the intersection at the same time, the vehicle to your right has the – you guessed it – right-of-way.  You have to wait for it to go through the corner.  Also, if two cars proceeding in the opposite direction stop at the same time, they may proceed through the intersection together – providing they are not turning.  If one is turning left, then that driver must yield to the driver going straight.  Still pretty simple, right?  It’s not.  Suppose you stop at the sign and notice another car to your left stops at the same time.  You’ll go first, right?  Maybe.  But if the other car is a low rider black Civic with tinted windows, including the driver’s window (a little illegal) so you can’t see that driver, you just might pause and let the bangers through.  At the other extreme, suppose the other car is a 1990s Eldorado with a sweet-looking little old lady at the wheel (she can barely see over it), very likely over-medicated, and in control of more automobile tonnage than God ever intended should happen.  You may want to give her a pass. 

    Now we go head-to-head.  You are stopped at an intersection with your left-turn blinker going, and across from you another car has stopped at the same time, with another car right behind it.  Of course, you let the oncoming traffic through.  Then it’s your turn, so you begin to execute your left turn.  But, whoa!  You have to stand on the brakes because the car behind the first car blows right through the intersection.  As you (maybe – depends on a lot of things) lean on the horn, you notice that the driver whizzing by is Hispanic.  He may not have read and memorized chapter and verse of the state of California Driver’s Manual.  For him, a different code applies:  He’s going straight; you’re turning left; you have to wait for him.  Another wrinkle here: Suppose you’re going straight and the other driver also starts through the intersection, then hangs a left in front of you.  More braking and horn-honking.  He didn’t have his left-turn blinker going – a device that strikes many southern California drivers (lots of them Hispanic) as a needless frivolity.  After you’ve driven those roads for a number of years and had innumerable encounters at four-way stops, you know that you must scan other cars and drivers as closely as you can, looking for a host of tell-tale indications of what might happen in the next few seconds. 

    When I see on TV breathless announcements of “the car of the future” that will drive itself, I like to imagine possible scenarios of that car encountering the drivers of southern California at that region’s hundreds of thousands of four-way stops.  The computerized car will “know” which car stopped first and which has the right-of-way according to the state Driver’s Manual, but – word to the wise – don’t hitch a ride in that car!  You’re likely to get yourself seriously injured.  Or worse – there are lots of those black Civic low riders cruising the streets, looking for trouble. 

-------------------

    Now, this little homily is meant to illustrate the complexity of seemingly simple communication.  I think it falls within the province of Michael’s HSR.  I chose the example purposely because, in another context and without my embellishments, the four-way stop is a textbook case of an elementary sign system in operation, of communication at work.  There are literal signs that, like Barry Sadler’s immortal Green Berets, mean just what they say.  S-T-O-P.  And there are unambiguous rules for the interpretation of those unambiguous signs.  S-T-O-P means stop!  And yet, life – that messy turbulence of incidentals that matter -- intrudes on our textbook example.  Bangers with homicidal intent, grannies zoning on Xanax, Hispanics just trying to survive north of the border, all distort that tidy picture of the communication-of-signs beyond recognition.  The California roll, in its culinary manifestation, is a bland, poor excuse for sushi, but dose it with enough soy sauce and slather it with wasabi, and it’s not half-bad.  Then it’s like life.    

Lee, great stuff. Goes nicely with Huon's mention of traffic in Beijing. But I'd like to challenge your conclusion, "all distort that tidy picture of the communication-of-signs beyond recognition." I would say that the tidy picture frames the scene within which the games are played. My inspiration is Pierre Bourdieu's frequent mentions of football (which Americans call soccer) as a model of social life. Habitus is neither the rules that define the shape of the field and the nature of the game. Nor is it routine behavior. It is the "structuring structures" that make it possible for a great player to kick a winning goal, having outmaneuvered the opposition. That the game contains many surprises does not mean that the rules are distorted beyond recognition. The rules define the space within which surprises occur.

RSS

Translate

OAC Press

@OpenAnthCoop

Events

© 2020   Created by Keith Hart.   Powered by

Badges  |  Report an Issue  |  Terms of Service