Source
Automatically imported from: http://commons.somewhere.com/rre/2000/RRE.Retrofit.and.Our.Tec.html
Content
| | | | --- | --- | | Red Rock Eater Digest | Most Recent Article: Tue, 15 Aug 2000 |
``` [Ben Sims just finished his PhD in the Science Studies Program at UCSD. His dissertation concerned California's program to retrofit its earthquake-prone bridges. I've enclosed (with his kind permission) the full text of his conclusion. It's a lucid case study in the ways that technologies are embedded in their social context, and you should read it. When he says "bridge", I want you to think "computer", and when he says "earthquake retrofit", I want you to think "Y2K retrofit". People like Rob Kling have explained for years just how intertwined a computer and its software can become with the customs and practices and strategies and knowledge around it, and how it can be enormously difficult to install new computers or even upgrade software as a result. Ben's discussion of the Coronado Bridge in San Diego, which is after all the sort of thing we used to mean by "infrastructure", throws an interesting light back on the infrastructures that we choose to focus on now. This being the conclusion, he does presupposes an understanding of some theoretical concepts that he has developed in earlier chapters. But if you push ahead you'll find that it stands pretty well on its own. I've heavily reformatted Ben's text and eliminated footnotes, references, and figures.]
---
This message was forwarded through the Red Rock Eater News Service (RRE). You are welcome to send the message along to others but please do not use the "redirect" option. For information about RRE, including instructions for (un)subscribing, see http://dlis.gseis.ucla.edu/people/pagre/rre.html
---
Date: Thu, 27 Jul 2000 22:18:43 -0600
From: Benjamin Sims
On Shifting Ground: Earthquake Retrofit and Engineering Culture in California
Abstract
Examines the history and organizational context of the seismic retrofit program at the California Department of Transportation ("Caltrans"), an ongoing effort to reinforce older freeway structures throughout the state to current seismic safety standards. In particular, focuses on how earthquake risks are understood in public and political contexts, how engineers assess the difficult-to-define risks posed to structures by earthquakes, and how these risk assessments enter into the design process. Also examines peer review in engineering and how Caltrans engineers and managers have coped with rapid changes in their knowledge base. The dissertation is based on archival and ethnographic research and interviews carried out at Caltrans, UCSD, and engineering design firms.
Table of Contents:
1. Introduction 2. Constructing Risk at Caltrans 3. Going Public: Engineering, Media, and the State 4. Civil Engineers: Peer Review and Professionalism 5. Change and Formalism in Design Practice 6. A Chain of Practices: From Laboratory to Design Floor 7. Conclusion: Retrofit and Our Technological Inheritance
Conclusion: Retrofit and Our Technological Inheritance
Section 1: Professional practice, standards, politics and personal interaction
Projection, as described in the previous chapter, is all about making connections between situations that are distant from one another so that the experience gained in one setting becomes useful in another. Though raised in the context of testing, this issue is really central to all kinds of work we call technical, and some we usually don't. Extraordinarily complex engineering and organizational projects must be built on an extensive division of labor among scientists, engineers, and skilled workers, each of whom may have a very narrow area of expertise. Typically, these groups also work in very different settings that do not overlap very much. Because these work settings are so different, each group develops a distinct local culture which may lead them to very different interpretations of the concepts and the artifacts they all work with. When politicians, the media, and local communities get involved, even greater cultural differences come into play.
Symbolic representations, such as written documents or computer programs, and more tangible objects, such as laboratory equipment or test specimens, frequently play a role in bridging these cultural gaps. But close examination of a social world like earthquake engineering reveals that the transfer of knowledge and the coordination of work between settings ultimately depends on extremely complex networks of personal interactions that cut across social boundaries. Despite the current proliferation of information and communications technologies, much of this interaction takes place face-to-face, perhaps necessarily so. Because face-to-face communication is very flexible and often seems to build trust between participants, it makes it easier to convey information between settings even if there is no consensus or even explicit discussion about what makes two situations similar. Things like ideas, practices, and test results may end up taking on quite different meanings in the translation, but enough cultural continuity can be maintained to permit coordinated action. In fact, the diverse interpretations may serve a purpose by permitting groups to analyze a problem from perspectives better adapted to the nature of their work. This emphasis on the coordination of work within and across social settings through chains of personal interactions came out in various ways throughout the thesis.
For example, one of the key factors that shaped definitions of seismic risk at Caltrans over time was the expansion of the "risk community" to include university researchers. Even though Caltrans engineers had read the academic literature before, and it had influenced their views, actually interacting with the researchers in person had a dramatic impact on the the way seismic risk was defined. This social event had a much greater impact on Caltrans definitions of seismic risk than the 1989 Loma Prieta earthquake itself.
In general, increased interactions between Caltrans engineers and outside peer reviewers brought out latent tensions between different segments of the profession. This is probably often the case when different social worlds come together, and in each situation the participants have to develop ways of managing these tensions if they are to successfully work together. In many peer review situations in engineering, participants try to minimize tension by acting according to norms of civility and disinterestedness. Peer reviewers also seek to avoid the appearance of competing with their colleagues for business by making sure they do not in any way seem to be usurping the role of the designers or questioning their competence. In the Caltrans case, academic members of peer review panels were able to cross this boundary freely, in part because they were not perceived as having commercial interests.
Any large engineering organization requires some means of regulating the design process if work is to be coordinated and standardized. As in the interaction between distinct segments of social worlds, both formal representations and personal interactions are important for this coordination, though symbolic representations can rarely get the job done by themselves. In the design context, these representations usually take the form of codes or other similar documents. Codes and less formal personal communications have very different capacities to shape design practice, capacities which give them distinct roles in the regulation of the design task. Codes can provide broadly accepted minimum design standards, but they are slow to adapt to changes in practice. On the other hand, individual people can assimilate new design approaches relatively quickly and pass their understanding along to colleagues, though discontinuities in the network of personal interactions can mean that some designers never adopt methods that are promoted in this way. In situations where practice is changing quickly -- here, as a result of increased interaction between designers and the academic community -- codes become increasingly irrelevant to the state of the art in design practice, and designers tend to rely more and more on their interactions with certain people within the organization who have access to the latest information and can communicate it effectively.
Chains of personal interaction aren't just useful for coordinating work, however: they may also serve as means for coordinating the production of authoritative interpretations of events, making them a tool for the consolidation of power, professional and otherwise. For example, media interpretations of the causes of structural damage in the Loma Prieta earthquake were heavily influenced by the views of the engineering profession. This was partly because engineers had already developed a more coherent view of what happened than any other group, and partly because, as credentialed experts, they fit the profile of reliable journalistic sources. But the media initially relied on political figures for interpretations of the events, because these were the sources they were familiar with. A crucial factor in bringing media accounts around to the view of the profession was Caltrans engineers' deliberate efforts to make themselves personally known and available to reporters. Though this exposed some disagreements in the organization, it was a dramatically successful strategy which had a decisive influence on public debate.
The engineering profession had a similar influence on the political response to the Loma Prieta earthquake. After setting a few ground rules, state government turned the investigation over to a board of inquiry composed almost entirely of engineers. Government agencies are generally comfortable turning these kinds of political-technical problems over to panels of researchers and professionals, because they give like-minded experts a chance to negotiate and reach consensus on whatever differences they may have. This provides a form of public accountability while avoiding the divisive debates between experts that often emerge in more public forums. Government officials and experts both gain a means for protecting and possibly consolidating their power to shape interpretations of events.
Finally, examining the process of testing and the use of test results in other social contexts demonstrates that the local skills and knowledge of one setting -- in this case, the laboratory -- can travel very far indeed through chains of representations, objects, and personal interactions that link work settings and social worlds. This provides a general view of how work and interpretations are coordinated within and between social worlds. Modern scientific research and technical work raise these characteristic issues in a particularly dramatic way because they depend upon extremely complex divisions of labor, but also seek to maintain a certain epistemological coherence. They therefore build up very complex and well-organized chains of interactions that support the widespread dissemination of knowledge and new technological artifacts. Though it is easy enough to focus on the global spatial scope of these interactions and the ability of modern scientific and technological institutions to transcend location, the crucial characteristic of these institutions is that the global connections they make are embedded, at nearly every point, in local, interactive circumstances.
Returning to technology
As this brief overview of the thesis suggests, it has largely focused on the nature and organization of technical work, and on the effects of rapid social and intellectual change on engineering practice. Seismic retrofit has been considered as one source of this rapid change. However, the very idea of retrofit also raises important questions about the nature of technology itself and the mechanisms that produce technological change. In order to give retrofit the attention it is due, in this chapter I return to the more traditional subject matter of the social study of technology: the nature of the interaction between humans and technology, and the question of whether we control technological change or it controls us. Examination of civil engineering work, particularly as it relates to the retrofit of existing structures, makes it difficult to sustain a view that we freely choose what path technological development will take under all circumstances. Neither does it support any kind of sweeping technological determinism. In the course of addressing these issues, this concluding chapter provides further evidence of the continuing importance of local circumstances and interactions in modern technical work.
Section 2: Engineering within the grid
An engineer designing a telephone or an automobile usually has a great deal of creative flexibility. Besides the technical standards of engineering practice, the forms of these kinds of objects are limited mainly by generic interface requirements -- a telephone should be able to work with standard phone lines, a car should fit within a standard traffic lane, etc. -- and by the availability of parts and materials. Working within these relatively broad constraints, the designer is free to give the object a wide range of different forms, depending on the anticipated wants of the purchasing public. Civil engineers work in a more limiting environment because every bridge or building they design has to fit uniquely into a particular local landscape, taking into account the existing infrastructure and geographical and geological features. But in either case, the designer is still working mainly with abstract representations stored in a computer or drawn on paper. Through these relatively flexible representations, a technological artifact presents itself to the designer as a malleable entity, changeable with a few strokes of a pencil or clicks of a mouse. There is a sense that the designer is in control of the form of the object, able to experiment with any number of variations to come up with one that seems to respond most elegantly to the constraints at hand.
The design of new technological artifacts is the stereotypical engineering task, and both engineers and sociologists have tended to treat it as definitive of engineering practice. However, many -- perhaps most -- engineers aren't designers. Some are researchers or managers, but the group that is most often neglected are those in charge of maintaining existing technology. For consumer goods like copiers or automobiles, such work is usually carried out by technicians or mechanics, but for buildings or bridges or sewer systems, the responsibility often falls on civil engineers. Maintenance engineers, like technicians and repairmen, and unlike most designers, work with technology after it has taken material form. They don't experience technological artifacts primarily as flexible representations, as designers do, but as relatively inflexible material objects: design is replaced by working on what is already there. As architectural critic Stewart Brand notes in the context of building maintenance and renovation, the best they can hope for is a "compromise with the fait accompli" of the object.
In civil engineering, however, designers are increasingly being forced to look at structures from a position similar to that of the maintenance engineer. This is a relatively new phenomenon in North America, particularly in the west, where the continual expansion of population and infrastructure had until recently put most designers to work on brand new structures in areas that had not been extensively developed, culminating in the grandiose freeway, aqueduct, and dam-building projects of the mid-20th century. As many areas have become more developed, however, engineers have gradually been forced to proceed in an atmosphere of constraint rather than limitless expansion. Environmental, historical, and cultural preservation laws passed since the 1970s have been one factor in this shift, as have newly restricted public works budgets. An equally important factor is that many projects have to be carried out within a matrix of existing infrastructure. The largest projects now tend to be things like replacements of aging freeway interchanges. Such a project requires engineers to take into account everything that has happened around the interchange since it was built, such as the growth of population and business in an area (which brings more drivers who depend on that interchange) and commercial development of land near the freeway or even right under it that was once open.
The emergence of retrofit as a category of civil engineering activity is, in many ways, a culmination of these trends. The choice to retrofit existing structures that have been deemed inadequate implies that it would somehow be too costly to simply get rid of them or replace them. In the Caltrans case, eliminating certain bridges would have been costly because they are part of a large freeway system that is the central means of transportation in the state. The state was not willing to bear the costs of replacing all of the deficient bridges that had been built during the rapid expansion of the freeway system in the 1950s and 60s; the kind of funding that supported this expansion was simply no longer available, and in any case the political establishment didn't want to wait 20 years for a fix. Finally, it would have been costly to abandon the old bridges and build new routes, due to the cost of acquiring land that had already been developed, and given current environmental and cultural preservation rules. Seismic retrofit is an effort to make dramatic changes in the transportation infrastructure while avoiding these excessive costs.
The constraints that face designers of new structures are intensified dramatically in retrofit work because the designers now must operate not only within the limitations imposed by surrounding infrastructure, but also within those imposed by the particular structure they are retrofitting. This means that designers have to find ways of learning about the structure as it exists in the field, not just on plans; they have to develop a personal knowledge of it, much as maintenance engineers do. Instead of working with an object that is flexible and relatively easily shaped according to their intentions, designers are confronted with an object that has most likely never been under their control and often seems quite capable of frustrating their intentions. The very constrained nature of this work is evident from the observation of one Caltrans engineer that retrofit projects require significantly more effort than designing a new bridge, but the end result (for a common overpass) might be only three sheets of plans, compared to about fifteen for a new bridge. The bulk of the effort is in documenting and analyzing the structure that is already there. Because designers are not used to working with existing structures in this way, they are acutely aware of the ways in which a structure that already has material form limits their creative options. For this reason, retrofit is an ideal site for both engineers and social scientists to come to a better understanding of the circumstances in which technology limits human action rather than expressing our intentions and giving material form to our interpretations of the world.
Section 3: Theories of technological change
Questions about how much technology can shape human action, or human action shape technology, are central to many theories of technological change. The two most important strands in Western technological thought take opposing views on the matter. What philosopher Langdon Winner has called the "voluntarist" view sees technological change as driven very straightforwardly by the rational application of technical means to solve objective human problems. Change is driven by the emergence of new problems and by the continual refinement of the available technical means. The path that technological development takes is the direct outcome of informed human choices. The other important theory, technological determinism, turns the voluntarist approach on its head. It holds that technology evolves according to an inner logic, not as a set of rational solutions to problems we identify. In fact, it suggests, human decisions ultimately have very little impact on the overall course of technological development. But technology, resistant as it is to human control, has an enormous determining influence on human affairs.
The voluntarist approach is significant because it suggests that technology should serve human needs, and that we are responsible for making the right choices about technology. But it falters in suggesting that the right solution to a problem can always be rationally identified, given the appropriate level of engineering expertise. The value of technological determinism is that it makes us aware that we are not always able to freely choose what shape technology should take next. It reminds us that the technology we develop today may have social consequences that cannot be completely controlled, even if they run counter to our future interests.
These views of technology, divergent as they are, share a common notion that the path of technological change can be comprehensively explained in terms of objective, rational principles. In the voluntarist view, new technologies follow directly from human needs: the problems to be solved can be clearly and objectively defined, and possible solutions are determined by fundamental scientific and engineering principles. In strong versions of technological determinism, technology evolves over time according to universal rules that can be identified by examining the historical record, and may be based in part on fundamental physical laws. In both theories, new technology emerges as the logical consequence of a given set of objective conditions.
As a result, neither has much to say about the activity of design itself. Because its results are explained largely in terms of objective forces, the particulars are presumed to be of little importance. This focus on objective forces also means that the design process is seen as largely free of social and cultural influences, even where society supplies the problem to be solved. But close examination of the work of engineering and invention, as a participant or as an observer, usually reveals that great creative effort is necessary to solve technical problems; that institutional arrangements and personal interactions can play important roles in the final outcome; and that engineers are not entirely insulated from prevailing cultural norms and assumptions. Looking at these circumstances, it is difficult to accept that new technology simply follows logically from well-defined social needs or existing technology.
A third, more recent view on the nature of technological change takes a social constructivist approach. Rejecting the idea that objective technological principles determine the outcome of the design process, social constructivists draw our attention to the "interpretative flexibility" of technology. They argue that different social groups attach radically different meanings to a single artifact, even to the point of having conflicting views about its basic operating principles. These attributed meanings can lead to different definitions of a problem and to radically different solutions. Though it retains the voluntarist sense that technology can be adapted to serve a variety of human needs can be responsive to human needs, the constructivist approach suggests that the flexibility of the meanings we attribute to technology is just as important as the flexibility of technology itself in shaping technological change. These meanings take shape differently according to the cultural dynamics of different social groups, making the evolution of technology an irreducibly social process. From this perspective, constructivism firmly dismisses determinist claims that technology develops according to an inner logic and has an independent influence on human affairs.
Social constructivism and interpretative flexibility
The view that technology is both materially and interpretatively flexible is the basis for a particular model of technological change put forward in the work of sociologists Wiebe Bijker and Trevor Pinch. They describe technology as developing through an ongoing process in which an artifact is invented; different social groups attach different meanings to it; these meanings lead to variations on the original form; and these variations are in turn interpreted by different groups, leading to further variations. The cycle usually ends eventually when "closure" is reached -- when an artifact develops a stable meaning across the relevant social groups, and hence takes on a stable form.
Bijker and Pinch use the history of the bicycle as an example. The first pedal bicycles, what were later called "penny-farthings", had pedals attached directly to a huge front wheel, with a tiny wheel in the rear for stability. Certain social groups, most notably young men, saw this bicycle as a "macho machine". They wanted a large wheel for greater speed, and didn't mind the bike's height and tendency to throw the rider over the handlebars. Other social groups who weren't interested so much in speed, like women and elderly men, saw the penny-farthing primarily as an "unsafe machine". The first interpretation led to the development of bicycles with ever larger front wheels, while the second interpretation inspired designs with lower front wheels and air tires, more similar to the modern bicycle. Eventually, "macho" riders saw that bicycles with low wheels and air tires could be even faster than penny-farthings. One form of the technology now satisfied both the macho need for speed and the need for safety. Closure had been reached and the artifact became stabilized in something like its modern form.
This basic approach has been elaborated by Bijker, Pinch, and others. Bijker expands the bicycle example and uses additional case studies of Bakelite and the fluorescent light bulb to paint a more nuanced picture of the institutional and industrial settings in which technology is developed. He introduces the concept of a "technological frame" -- a dominant interpretation of a technology, usually backed up by certain tools and techniques -- arguing that certain actors are included in a particular frame more than others, which explains why, for example, engineers and marketing executives have more influence over the design of fluorescent light bulbs than the average consumer. Donald MacKenzie draws indirectly on Bijker and Pinch in his study of the history of nuclear missile guidance systems. He argues that institutions, rather than internal principles of technological development, shape the trajectory of technological change, and that different social groups -- such as opposing pro-bomber and pro-ICBM groups within the military -- developed radically different interpretations of missile tests and accuracy estimates. Ronald Kline and Trevor Pinch show how users can reinterpret and modify already-stable technological artifacts, as when farmers modified early automobiles to power saws and washing machines or to serve as tractors.
Interpretative flexibility and the context of invention
All of this work emphasizes the extreme flexibility of technology. Little attention is paid to the ways in which technology may sometimes frustrate our efforts to control it and constrain our actions. If a technology seems to be resistant to certain lines of development, this is usually explained in terms of social or institutional stabilization. Furthermore, Bijker and Pinch explicitly argue that the material flexibility of technology is an integral part of the idea of interpretative flexibility, and the other authors all seem to assume this to some extent. The work of invention and design appears to lie almost entirely in trying to understanding the interpretations and satisfy the needs expressed by various social groups, while the form of technology follows easily from these interpretative efforts.
This vision of technology makes sense mainly because all of these authors have chosen to look at a particular aspect of technological change: design and invention, particularly in reference to relatively inexpensive mass-produced consumer goods and weapons. These are usually replaced frequently over time -- middle-class Americans keep buying new cars, the military continually procures new weapons. Because they are looking at these kinds of technologies, when these authors talk about different interpretations of an artifact -- the bicycle, for example, or even one specific type of bicycle -- they are not really talking about any particular artifact, but rather a succession of different artifacts, each copied many times over. No wonder the meanings we attach to technology appear to be so readily translated into material form in their work: the changes they describe do not involve modifying existing objects, but turning concepts into brand new products that will be built from basic parts and materials.
Most technological work is not nearly so open-ended. Civil engineers, I have argued, face a much more constrained design task, especially when they are engaged in tasks like seismic retrofit. In their work, technological artifacts often appear as independent entities, and it sometimes seems to take a great deal of engineering effort to bring them back under control. In order to account for this, the constructivist approach should not be so quick to throw out all vestiges of technological determinism. Specifically, it ought not to assume that interpretative flexibility entails material flexibility as well. These should be treated as two separate concepts, one referring only to the idea that people can attribute many different meanings to a given artifact, and the other referring to the extent to which we are able translate our interpretations into material form. We need to look more closely at circumstances in which technology doesn't seem to be as flexible as our interpretations of it. Seismic retrofit is one such circumstance.
Section 4: Engineers confront bridges
There are at least three ways that engineers encounter objects -- in this case, massive civil structures like bridges -- as independent, inflexible entities through retrofit work: structures are durable in the face of changing engineering practice, they change over time in ways that can't be easily documented, and they become embedded in local material and social settings.
Durability and change
The principle of interpretative flexibility, so long as it is divorced from material flexibility, applies as much to buildings and bridges as to bicycles and light bulbs, and as much to the engineering profession as to any other social group. Principles of good engineering are constantly being updated in the light of experience and research. But when a new generation of bridges appears, we can't just throw the old ones away in landfills and junkyards, as we would with many consumer goods. They are too expensive and too crucial to a wide range of activities for that. So regardless of how much engineering practice has changed, relics of past practice will always remain, even though they may come to be considered outdated or defective. Bridges are durable even though engineers' interpretations of them are not.
There is no better example of rapid change in civil engineering practice than in the area of earthquake engineering during the 1970s and 80s. In the 1960s, engineers designed bridges in California with widely-spaced hoop reinforcement and unrestrained deck joints, checking only to see if they would stand up to a small horizontal force; at the same time, they used the latest computer tools to design bridges that soared more gracefully and had fewer support columns than before. At the time, of course, these bridges were seen as symbols of cutting-edge engineering and as models of seismic safety. Then in the 1970s, 80s and 90s, the specialty of earthquake engineering expanded rapidly and found new sources of funding, leading to more research and a much more detailed understanding of how structures respond to earthquakes; these new understandings were incorporated into more and more sophisticated computer programs made possible by increasingly fast computers; and, finally, major earthquakes caused damage to Caltrans bridges and provided the impetus for engineers to look again at the structures they and their predecessors had designed before. The definition of an earthquake-safe structure changed dramatically: columns should be designed with closely-spaced spiral reinforcement, joints should be restrained so they won't move too far, and underlying seismic risks should be analyzed more carefully.
It was through these changing interpretations of earthquake risk and structural performance that many of the older bridges in California came to be seen as unsafe. Like any engineers faced with such a situation, Caltrans bridge designers responded to these new interpretations by building new bridges differently. But they also began to consider ways in which the existing bridges could be made safer through retrofit. This task took on added urgency when geological and political circumstances combined to draw public attention to the fact that many bridges did not measure up to current engineering standards. But while designers can decide to do things differently with new bridges, they aren't able to simply forget the past while doing retrofit work. Instead, they are forced to grapple directly with the engineering practices and design decisions of their predecessors. Even engineers who would not ordinarily be interested in the history of their field get the opportunity to learn something about it through this work. Technology takes on an element of autonomy here simply because it can't change in response to our new ideas about it.
Documenting change
Structures also escape the control of engineers over time by changing, or being changed, in ways that are not or cannot be documented in the visual language of engineering plans. Engineers and architects try to maintain true representations of structures by retaining and updating what are commonly known as "as-builts": engineering drawings initially created during the construction process in order to document how a structure was actually built, including any deviations from the original plans. These, rather than design drawings, are usually retained on file for the life of the building or bridge. Theoretically, as-builts are supposed to be updated continuously to reflect modifications, but in reality they are often neglected. Brand notes that the as-builts for factories are usually scrupulously updated every time any change is made, simply because the building is one of the key assets of a manufacturing firm. But often design engineers and architects don't pay a lot of attention to as-builts, because most of the time they aren't directly concerned with what happens after something is built.
Caltrans keeps as-builts for all their bridges on file, but keeping them up-to-date was apparently not a top priority, as engineers found out when they needed to retrofit. Sometimes the as-builts did not accurately reflect what went on during construction. This often had to do with irregularities that the engineers who were on-site during construction might not have known about -- piles that ended up at not quite the right angle when they were driven into the ground, or footings that spread out into the surrounding soil more than usual when the concrete was poured. Sometimes what was done during construction was simply not documented accurately. For example, designers working on the retrofit of the San Diego-Coronado bridge, one of the state's major toll bridges, needed to know whether lateral stiffeners on some of the bridge's huge girders had been welded to the top flange of the girder or not in order to put together an accurate computer model. In this case, the detail was probably left out because it is conventional not to make such welds, but the designers had to be certain.
In other instances, significant changes made after construction were never recorded in as-builts, or were recorded in some sets of as-builts but not others. For example, designers working on retrofits after the 1989 earthquake on at least one occasion determined by looking at as-builts that a bridge's deck joints needed retrofitting, only to find out during construction that retrofit devices had already been installed. When I toured the San Diego-Coronado bridge with one retrofit designer, our tour guide, who worked in maintenance and was very familiar with the structure, pointed out a number of features that were not on the designer's as-builts, including electrical conduits that had been installed along the bottom of the bridge deck to supply power to a machine that moved lane dividers, and holes that had been drilled in the deck for the insertion of lane-marking poles and had caused the underlying steel to rust. These particular details weren't structurally significant, but could have interfered with the installation of retrofit devices.
Structures also change over time without any direct intervention from people -- from age, use, and exposure to the elements. Concrete, for example, tends to shrink with age, and other materials may deform in unexpected ways as well. Bridges have expansion joints and other movable parts that accommodate these changes. Caltrans maintenance engineers regularly inspect bridges and note any serious problems, but they don't usually make note of minor, expected changes, such as small movements at expansion joints. Inaccessible parts of structures, like underground or underwater piles, are inspected less regularly because of the effort involved, so less up-to-date information is available about them. Usually, any serious deterioration is noted before it poses a risk to the structural integrity of the bridge, but retrofit designers also need to know about less significant changes so they can properly model the structure.
Engineers cope with the uncertain relationship between as-builts and structures in the field in a number of different ways. One strategy is to dig through more archives -- maintenance records or files at the Caltrans district office in charge of the bridge, for example -- to try to get a more complete list of changes. Very often, though, the uncertainties are only fully resolved by direct inspection. For the Coronado Bridge retrofit, for example, divers were sent to document the condition of underwater concrete piles, and found cracking that had exposed some of the reinforcing steel, causing it to rust. In this case, models of the piles were being tested in a university laboratory, and the cracks were replicated on the models. The retrofit team also took measurements themselves, in one instance shutting down some lanes of the bridge for a morning so they could go out with tape measures and determine how far the expansion joints had separated. Even after such inspections, however, there are still some changes that only become apparent during the construction process, such as when workers try to put down new piles and find themselves hitting existing ones, or when they drill into the concrete to install seat extenders -- metal bars placed across an expansion joint to prevent it from separating -- and run into the bars that are already there.
These examples illustrate the difficulties engineers face in maintaining accurate representations of an object over time. In the case of Caltrans, it is true that engineering representations could have been made much more accurate in many respects simply by putting some effort into developing a more integrated and well-organized record-keeping system, as Brand suggests we do for buildings. But the rapid accumulation of changes on an enormous structure like the San Diego-Coronado bridge would still stretch the limits of such a system. More fundamentally, engineers will always have to use some kind of criteria to decide which changes are important, because documenting every tiny change in a large structure could be an infinitely time-consuming task. And since choices about what aspects to document will inevitably reflect the engineering culture of the time, there is no guarantee that the records that are kept will be adequate for the needs of future engineers. Engineers' representations of an existing structure will always be at least a little bit out of date and a little bit inaccurate. Because engineers' control over the material world depends largely on their skills in manipulating representations, this means that objects are always just out of the control of engineers once they are built. When retrofit becomes necessary, designers are forced to go out and learn directly from the structure.
Embedding and resistance
Structures can also change independently of the intentions of engineers by becoming embedded in local material or social circumstances over time. In the material sense, structures often end up interpenetrating with other elements of the local infrastructure. Bridges are sometimes used to carry electrical or communications lines, or even water pipes. Lighting and other services might be mounted on the bridge structure. Most of these changes are of the sort that could be documented by an agency like Caltrans, but in practice they do not always make it into as-builts because they are installed under the authority of other local or state agencies.
Bridges and other structures also become embedded in a similar way in local history and social circumstances. Here I'm not referring simply to their functional role in enabling transportation and changing mobility habits, though these are important. I am also referring to the cultural meanings that people in the wider community invest in material objects as a result of interpretative flexibility. These meanings can stabilize over time, out of the control of engineers, placing serious constraints upon them when it comes time to retrofit. Of course, bridges begin to acquire such meanings the moment they are conceived in the mind of some engineer, planner, or politician. Therefore, even the design of a new bridge is an act of "heterogeneous engineering" that requires engineers to take a wide range of actors and meanings into account. It is important to understand how this process works in order to see how the retrofit design process is different. Take the San Diego-Coronado Bay Bridge as an example.
Building the Coronado Bridge
Various groups made plans as far back as the 1920s to build a bridge between the city of San Diego and the "island" of Coronado, which is actually the end of a peninsula that connects to the mainland well south of San Diego via a narrow, sandy causeway. These efforts were supported mainly by the business community and those with an interest in expanding tourism, but in every case they failed either because of intense opposition from Coronado residents who saw it as a threat to the small-town atmosphere on their "island" or because of concerns raised by the Navy that a bridge would restrict access to the naval base in San Diego Harbor. Since many Coronado residents were retired military officers, the two interests reinforced one another. During the 1950s and 60s, planning for a bridge began in earnest with the support of Governor Edmund G. Brown, a champion of bridge-building as a tool of economic development. During the planning stage, engineers were forced to take Navy concerns into account by including underwater tunnels as an option in place of a bridge. In the end, possibly because of cost concerns, a bridge was chosen over tunnels, and design began in the early 1960s. The Navy softened its stance against a bridge, possibly because of political pressure from Brown allies in Washington.
Initial designs called for a straight bridge between San Diego and Coronado. With a 6% grade -- which is considered very steep for a bridge -- the bridge would have 180 feet of clearance in the center of the channel. The Navy, however, demanded at least 200 feet to accommodate its largest ships. The engineers' solution to this dilemma was to maintain the grade as it was but to make the bridge longer by putting in a 90 degree curve in the middle of the bay, making the added height possible.
There were other important influences on the final shape of the bridge. For example, the location of the two ends of the bridge was dictated in part by the need to place the highest portion of the bridge directly over the main shipping lane, which is closer to the San Diego side. The bridge structure therefore touches down at the very edge of Coronado, while extending perhaps a quarter of a mile inland on the San Diego side to meet up with Interstate 5. This was made possible, however, because the San Diego approach ran straight through a poor Mexican-American residential area. This was a "path of least resistance" at the time, where property values were low and where residents had little political voice. Having the bridge come only to the edge of Coronado took care of a potential reason for opposition to the bridge from politically connected Coronado residents who wanted minimal disruption of their lifestyle.
As this example shows, engineers balance a host of technical, political, and economic considerations as they design a bridge. Because the shape of a bridge is very flexible as long is it remains on paper, designers are able to accommodate the form of the bridge to satisfy these heterogeneous constraints. The layout of the Coronado Bridge, for example, represents in material form a compromise between the conflicting interpretations of the Navy, politicians, community groups, and engineers, though it deliberately ignores the interpretations of less powerful groups. After a bridge is built, however, it loses much of its material flexibility. As it stands, though, it continues to collect new interpretations, and becomes part of the history and social fabric of a community. High-profile bridges like the Coronado Bridge, and perhaps most famously the Golden Gate Bridge in San Francisco, become symbols of a community, both to residents and to the world. But bridge designers are no longer professionally concerned with these interpretations, having moved on to other projects, and may be unaware of the extent to which a structure has worked its way into the local culture. When they begin the task of retrofit design, engineers often have to confront a set of interpretations that have stabilized in their absence. There may be less room for negotiation and compromise with such entrenched interpretations.
Retrofitting a work of art
Engineers involved in the seismic retrofit of the Golden Gate Bridge and the replacement for the seismically-deficient San Francisco- Oakland Bay Bridge, for example, have had to carefully negotiate with a host of Bay Area planning and environmental commissions intent on preserving these bridges as city landmarks. The City of San Diego has far fewer of these groups, and the Coronado Bridge, while well-loved by many, has not taken on as much symbolic value as the San Francisco bridges. But it ironically developed a great deal of cultural significance to the one neighborhood it nearly destroyed: the largely Chicano community known as Barrio Logan, bisected by the San Diego approach to the bridge. Barrio Logan had been a thriving neighborhood enclave, but by the early 1960s was already suffering from zoning laws that were changed to allow industrial use and from the construction of Interstate 5, which cut it off from the larger Logan Heights neighborhood. The construction of the bridge accelerated these changes and nearly destroyed the neighborhood.
But by the late 1960s, Chicano political activism was on the rise nationally, and was becoming particularly important in California. In 1970, the Barrio Logan community, still upset about the building of the bridge, took action when they discovered that an enormous California Highway Patrol station was to be built on the already desolate empty land under the approach ramps to the bridge. The community was not happy with the prospect of such a heavy police presence, so several hundred residents occupied the construction site for twelve days, demanding that a community park be built instead. City officials intervened and granted the request, and what came to be known as "Chicano Park" was established under the approach ramps. This action is remembered by many as a pivotal moment in the developing political and cultural consciousness of Barrio Logan residents and of the broader Chicano community in San Diego.
Though the park was a focal point of community activity thereafter, it was a noisy and sometimes gloomy place, dominated by the gray concrete of the bridge structure. Partly in order to combat this gloominess, a loose coalition of local artists conceived of the idea of painting murals on the bridge columns in the park. In this, they were part of a larger revival of Chicano "muralism", inspired by Mexican artists like Diego Rivera, Jose Orozco, and David Siqueiros, that saw murals as an expression of community solidarity, cultural heritage, and political resistance. Over the next thirty years, no less than forty brightly-colored, symbolically-dense murals were painted on bridge columns and abutments in the park. Most were painted by local artists, but major muralists from throughout California and the southwest contributed as well. Some murals depict cultural figures and events, or display political messages like "VARRIO SI, YONKES NO" (barrio yes, junkyards no) or "NO RETROFITTING". Others make more purely artistic statements, such as one titled "Collosus" that depicts a muscular figure carrying the bridge deck on his back, his torso painted on the column and his arms outstretched along the beam supporting the deck. Another, "Tres Grandes y Frida", is a striking, impressionistic portrait of muralists Orozco, Rivera, and Siqueiros with painter Frida Kahlo. These examples only hint at the diversity of themes represented in the murals in the park.
As painting continued, the murals came to represent a certain degree of community ownership of the bridge that had so damaged the neighborhood. And to some, the bridge itself began to take on new aesthetic qualities. To many people in San Diego, the most striking feature of the bridge is the curving, bright-blue sweep of the deck girders as the bridge extends across the bay. But from Chicano Park, the most striking feature is the pairs of columns marching in a line toward the bay, oddly resembling the nave of a cathedral. This isn't completely coincidental, since the architects who consulted on the design of the bridge had intentionally designed the columns to resemble mission-style arches, a fact which is best appreciated from under the bridge. The depth of this reinterpretation of the bridge by the artists and the community is reflected in comments made by muralist and community activist Salvador Torres to a newspaper reporter in 1989:
When I look into the depth of the columns, as the arches flow toward the waterfront, I hear a sound, a mystical sound, like that of a living creature. To me, the bridge is life -- reassurance, reaffirmation ... love. I know what the birds must feel when they fly over. They feel pride, in a fortress of beauty and strength.
While some Caltrans engineers and some of the local bridge engineers that were contracted to design the retrofit were aware that the murals could pose a problem, they do not appear to have been aware of the depth of the community's commitment to the artwork. Early in the design process Caltrans held public meetings where it presented a full range of possible retrofit measures, most of which would have had a significant impact on the murals -- including completely replacing the columns, encasing them with steel jackets, and thickening the existing columns in the lateral direction. Both community activists and elected officials began to protest immediately. Decades of distrust between the community and Caltrans rose to the surface, and some questioned whether the bridge really needed to be retrofitted at all. An artists' group sent out a newsletter demanding "no retrofit, not now, not ever!" Local politicians and newspapers soon took up the cause as well. Clearly, the power structure in the city had changed considerably since the 1960s.
Caltrans engineers and design consultants were in a difficult position: they were pretty sure they would have to retrofit the columns in a way that would affect the murals, yet this did not seem to be a viable option, politically speaking. Caltrans officials in Sacramento had meanwhile appointed a peer review panel to oversee the retrofit design. One member of this panel was Frieder Seible, structural engineering professor at U.C. San Diego. By this point, the Caltrans local district personnel who were in charge of community relations had realized they could not handle the situation on their own in the prevailing atmosphere of distrust. So they asked Seible to intervene, with the idea that he could explain the necessity of retrofitting to the community. Seible invited community activists to tour the UCSD structural engineering laboratory, where he explained to them the basic reasons for retrofit and showed them test specimens that had been put through simulated earthquakes. This made a big impression, and convinced many of those present of the need for retrofit. It was apparent to Seible that the community trusted him much more than Caltrans officials, perhaps because of his academic position and perceived independence from the department.
In addition to convincing the community that retrofit was necessary, Seible also pressured project designers to do a very detailed analysis of the columns, instead of relying on standard Caltrans design methods and retrofit techniques. Something like steel jackets around the columns would be standard in this case, because there were "lap splices" in the reinforcing steel. What this means is that each vertical reinforcing bar was not a continuous piece of steel from the foundation through to the cap. Instead, each strand of vertical reinforcement consisted of two bars that overlapped for a few feet within the column, held together only by the surrounding concrete. This was standard practice in the 1960s, but engineers have since come to understand that lap splices can easily pull apart in an earthquake. This can be avoided by "clamping" the lap splices more firmly in the concrete, for example by placing a steel jacket around the column. In this case the lap splices had been staggered, so that some were in the middle of the columns, while others were very close to the footings, below ground level. Based on the more detailed calculations and new soil data, it was determined that the columns could be strengthened sufficiently simply by adding a new "mat" of reinforcing steel and concrete to the top of the footings, which would strengthen them and at the same time provide confinement for the lowest set of lap splices. It was also necessary to do some work near the top of the columns. But since most of the work would be done below ground level, the murals could be preserved. The retrofit designers could feel more comfortable taking this unconventional approach because on the peer review panel, Seible had main responsibility for the Coronado retrofit, and he had already endorsed the approach.
Design under pressure
The case of the Coronado bridge retrofit and the Chicano Park murals is a vivid illustration of how the factors discussed above -- the durability of structures in the face of cultural change, the undocumented changes that occur in them over time, and the way the become embedded in local material and social circumstances -- can come together to constrain the designer's task. The murals, in particular, combine the social and material aspects of embedding; the bridge developed great significance to the community not only because of the social interactions that occurred around it, but through their material transformation of the structure and the land surrounding it. Caltrans engineers and officials faced a social and political problem that could ultimately be solved only within the material form of the retrofitted bridge. The murals were also a change that had been made to the structure outside of the control of engineers, a change which they had to assimilate into their practice. Finally, designers had to see how the columns had been built originally, and figure out how to analyze their outdated reinforcement configuration, which further restricted their range of action. After putting in so much engineering effort, and after all of the complex negotiations with the community, the end result was a fairly straightforward footing retrofit, along with some changes at the top of the columns. After working their way through all of the constraints, when the engineers got to the point of actually designing the reinforcement mat for the footing, there was little flexibility remaining. Most of the effort went into learning about the material and social structures that existed in the field and had long ago escaped the control of design engineers. Some of the project engineers felt they had been lucky to find any engineering solution to the problem at all.
Section 5: Autonomous technology and human choice
When they work on retrofit projects, engineers find that even after they have addressed the concerns of all relevant social groups and have satisfied the technical conventions of their profession, the range of possible solutions to the design problem is limited still further by the existing material object that is the focus of the retrofit effort. This is not to say that technology has an objective, internal logic that it unambiguously imposes on engineers, or that it possesses a human-like agency. As constructivist studies of technology have shown, engineers base their design efforts on their own, culturally-conditioned interpretations of the world. They interpret the needs and desires of the relevant social groups, the conventions of professional practice in their field, and the relevant physical principles, and each of these interpretations further constrains the design. When they are working with an existing structure, they bring it into the design process through socially-mediated interpretations, as well. They look at maintenance reports and as-builts; they go out to the structure and inspect and measure it. The understandings they take away from these activities are always filtered through their professional experience and socialization. Still, just having another object to consider ultimately restricts their options even further. Through their interpretative activity, the existing structure has a causal impact on the final design. And through the medium of the structure, the designers are also grappling with specific engineering decisions made by their predecessors.
Our technological inheritance
Though the sort of technological determinism that claims there is a logically determined sequence to technological change is not supported by this analysis, another, more ironic, form of technological determinism is. This form of technological determinism is best explained in reference to large, durable objects like civil engineering structures, which are too expensive to simply get rid of or ignore. Given that an existing structure does seem to limit the range of actions we are willing to take, and given that many structures that were built, say, 30 years ago are still around, the decisions that engineers made 30 years ago, based on their interpretations of technology at the time, have an effect on the decisions we make today, even though our interpretations of technology have changed. In turn, our decisions, limited as they are by those taken 30 years ago, are limiting to future decisionmakers. Barring some unprecedented disaster, we will never be given the luxury of starting with a completely clean slate. Even though technological change is ultimately driven by human choices, the durability of our past choices limits our future courses of action, and technological change develops momentum in certain directions in ways that we don't have complete control over.
Other researchers who have studied infrastructure have come to similar conclusions. Historian Thomas P. Hughes, whose work focuses on electrical power systems, argues that technological systems acquire a kind of momentum over time which predisposes them to evolve in certain directions. This is a result both of the inertia imposed by organizations and individuals that become committed to a particular direction of change, and of the durability of the material objects that make up a system, which are often too costly to replace in the short term. Similarly, Susan Leigh Star and Karen Ruhleder, in their study of information infrastructure, note that new infrastructure is never designed "de novo" but always "wrestles with the inertia of the installed base".
The study of infrastructure is important for understanding the nature of technological change because it draws attention to the fact that future generations will have to live with the technological choices that we make today, and won't be entirely free to ignore the mistakes we make. But it would be wrong to restrict this analysis to infrastructure. Even consumer goods with a limited useful life don't simply disappear into thin air when we are finished with them. Every insignificant plastic bag, lawn chair, or bicycle that has ever been made is still with us in one way or another, most likely taking up space in a landfill. Other technological byproducts, like radioactive and chemical waste, pose dangers that we have to address over time in a much more active way. Furthermore, goods that are assumed to have a limited useful life among the wealthier nations and social classes of the world are often far too expensive for others to simply discard; hence, people continue to maintain and drive older, more polluting cars even as we try to make new cars more and more environmentally friendly. Because technology is durable, we must always make new interpretations and choices amid the debris left by our ancestors.
Section 6: The local embedding of globalizing technologies
Though the emphasis here has been on ways that technology can sometimes escape our control, it has also been noted at every step that technological artifacts are surrounded by locally embedded social interactions -- interactions based on personal familiarity and face-to-face interaction. Engineers designing or retrofitting a bridge work in close-knit teams that are bound together by personal interactions, and the feelings of a community toward a bridge, for example, are articulated in the meetings of planning commissions and neighborhood activists. In fact, when a technological artifact is surrounded by embedded social circumstances in this way, it can be said to become locally embedded itself. People's interpretations come to be based not on abstractions or broad cultural symbolism, but on a kind of personal familiarity derived from direct contact with the artifact. Engineers confront a bridge in person in order to design a retrofit; a neighborhood makes an imposing bridge into a familiar focal point for social interaction. A bridge that increases people's geographical mobility, and hence further disembeds some social interactions, can simultaneously become part of a new set of embedded social relationships.
Anthony Giddens and others have argued that modernity is fundamentally about the dialectic between the local and the global. The disembedding brought on by increased globalization is said to open up new ways of attending to the local, a process Giddens calls reembedding. But his analysis of reembedding draws on examples like this:
The self-same processes that lead to the destruction of older city neighbourhoods and their replacement by towering office-blocks and skyscrapers often permit the gentrification of other areas and a recreation of locality.
By focusing on such self-conscious attempts to recreate a lost sense of locality, theorists of modernity may miss a more fundamental aspect of the relationship between the local and the global. The very mechanisms that make disembedding possible -- modern transportation or communications infrastructure, or social institutions like financial markets -- aren't just abstract structures. They exist and are given meaning only through the efforts of the groups of people who run and maintain them. They can't be built up without the simultaneous creation of locally embedded networks of interpersonal interaction. This chapter has shown that technology, in particular, is always already a product of embedded social relations and locally-generated interpretations. Focusing on the social impact of technology -- its potential to disembed and reembed social relations on a broad scale -- at the expense of the social context of its production is, ultimately, to promote an unwarranted form of technological determinism.
end ```
| | | --- | | ProcessTree Network TM For-pay Internet distributed processing. | | Advertising helps support hosting Red Rock Eater Digest @ The Commons. Advertisers are not associated with the list owner. If you have any comments about the advertising, please direct them to the Webmaster @ The Commons. |