memory practices in the sciences / geoffrey c. bowker

Geoffrey C. Bowker
1 Review
MIT Press, 2005 - Philosophy - 261 pages

Winner, 2007 Ludwig Fleck Prize given by the Society for Social Studies of Science (4S). and Awarded "Best Information Science Book 2006" by the American Society for Information Science and Technology (ASIS&T).

Google Books.

The way we record knowledge, and the web of technical, formal, and social practices that surrounds it, inevitably affects the knowledge that we record. The ways we hold knowledge about the past—in handwritten manuscripts, in printed books, in file folders, in databases—shape the kind of stories we tell about that past. In this lively and erudite look at the relation of our information infrastructures to our information, Geoffrey Bowker examines how, over the past two hundred years, information technology has converged with the nature and production of scientific knowledge. His story weaves a path between the social and political work of creating an explicit, indexical memory for science—the making of infrastructures—and the variety of ways we continually reconfigure, lose, and regain the past.

At a time when memory is so cheap and its recording is so protean, Bowker reminds us of the centrality of what and how we choose to forget. In Memory Practices in the Sciences he looks at three "memory epochs" of the nineteenth, twentieth, and twenty-first centuries and their particular reconstructions and reconfigurations of scientific knowledge. The nineteenth century's central science, geology, mapped both the social and the natural world into a single time package (despite apparent discontinuities), as, in a different way, did mid-twentieth-century cybernetics. Both, Bowker argues, packaged time in ways indexed by their information technologies to permit traffic between the social and natural worlds. Today's sciences of biodiversity, meanwhile, "database the world" in a way that excludes certain spaces, entities, and times. We use the tools of the present to look at the past, says Bowker; we project onto nature our modes of organizing our own affairs.

REVIEW ONE

Wilson, T.D. (2006). Review of: Bowker, Geoffrey C. Memory practices in the sciences. Cambridge, MA: MIT Press, 2005. Information Research, 12(1), review no. R238 [Available at: http://informationr.net/ir/reviews/revs238.html]

Geoffrey Bowker has had an interesting career, ending up (to this point at least) as Regis and Diane McKenna Professor at the University of Santa Clara after spells in Paris, at the Centre for the Sociology of Innovation, the Graduate School of Library and Information Science at the University of Illinois, and the Department of Communication in the University of California at San Diego. Probably his best known work (with Leigh Starr) is Sorting things out, a study of classification, and, to a degree, this work is in the same general area.

The term 'memory practices' is not immediately understandable—or, at least, it was not to me—and that might damage the books chances in the library and information sciences market place. That would be a pity, since this is the kind of book that can stimulate a dozen Ph.D. studies, or more, and that contributes a genuine intellectual dimension to the field, which is so often lacking in our technology oriented (one might say 'technology bound') age. Bowker's definition of 'memory practice' is presented almost tangentially in a section (p. 6-15) on What this book is about. He notes that,

Acts of committing to record (such as writing a scientific paper) do not occur in isolation; they are embedded within a range of practices (technical, formal, social) that I define as memory practices. Taken as a loosely articulated whole, these practices allow (to some extent) useful/interesting descriptions of the past to be carried forward into the future. (p. 7)
Earlier, he notes that the traces we leave of our actions are not truly the way they were, or the way they happened, but 'a tacit negotiation between ourselves and our imagined auditors': we write accounts of research not to truly report exactly what happened, but to observe the rules (formal and informal) of scholarly communication. A comment that reminds me of John Johnson's PhD thesis on The social construction of official information, in which he records how social agency statistics were reported so as to gain the biggest budget benefit. Bowker has a similarly interesting tale, from earlier research in the field of company history:

A similar move [to the current company policies of setting rules for e-mail communications] was made in the 1930s by the Schlumberger company, when it realized that its internal records could be scrutinized by a court—the company shifted very quickly from writing detailed accounts of their practices in French to writing highly sanitized versions in English. (p. 7)
Following the introduction, Bowker addresses 'memory practices' in three areas and eras: geology in the 1830s, cybernetics in the 1960s, and environmental sciences (with a focus on biodiversity) today.

In the first of these, geology itself is presented as a 'memory practice', the emerging knowledge of the earth's construction revealing its history, the strata themselves containing the records of ages past. Of course, as geology was developing so was the scientific literature in many fields. Bowker notes that geology's focus on the temporal distribution of the geologic record and the relationship of this, in the scientific mind to watchmaking and, from there, to mechanical models in general.

In the cybernetics chapter Bowker suggests that the 'memory practice' of the field was, in fact, the destruction of memory; all things being reproducible from first principles. Memory was seen by Ashby, for example, as something needed in man only because of our '...incomplete knowledge of the appropriate Markov chain for the given closed system'.

Finally, in Databasing the world: biodiversity and the 2000s, Bowker examines the present reliance upon the database, underpinning the information infrastructure built upon, among other things, the Internet, showing that, in historical terms, we have moved from perceiving the world as a clock (geology) to the world as a computer (the world of the Internet and the database). Today's technology leads to collaboration in science, to information sharing and information access beyond anything previously available. The database becomes memory in an analogic way to the rock strata becoming memory.

The three case chapters are followed by three further chapters: Chapter 4, The mnemonic deep: the importance of an unruly past explores the problems of naming things and classifying them and the problems arising from multiple classifications:

...we need databases held together through good metadata practice; and in order to work out the latter, we need to recognize the depth of the problem. In a biodiverse world, we need to be able to manipulate ontologically diverse data.
Chapter 5, The local knowledge of a globalizing ethnos attempts to answer the question, Why keep knowledge local and local knowledge? Bowker notes that the first answer is a simple one: knowledge is inevitably local. However, he does not suggest any other answers, although his following text leads one to suppose that his comments on different ways of knowing are intended to convey the notion that knowledge held locally will be viewed in different ways and, perhaps, lead to different kinds of understanding.

In the final chapter, the concluding paragraph effectively summarises what the book is about:

Just because the past is over doesn't mean that there is a truth about "wie es eigentlich gewesen ist". The work of creating partial objects and conjouring them into a given, small set of trajectories is a work in the present of expanding our empire and our knowledge. If we want the future to be other than what it seems to be turning out, we must create a past that is other than it seems to have turned out. People, planets, and purgatory (Le Goff 1984) deserve multiple pasts. Only an open past can unlock the present and free the future.
This is a very interesting work, but I think it could have benefited from the attentions of a good editor. The style is dense and somewhat quirky, with parenthetical interpolations in many sentences, that interrupt the flow of understanding—and not always usefully. In spite of this, I unreservedly recommend the book to anyone who seeks to understand how, in science, we can ensure that what is learnt in the past and the present is available for the future.

Professor T.D. Wilson
Publisher/Editor-in-Chief
September


REVIEW TWO

Sarita Albagli
[www.reciis.cict.fiocruz.br] ISSN 1981-6286

Pesquisadora do Instituto Brasileiro de Informação em Ciência e Tecnologia (IBICT/MCT), Rio de Janeiro, Brazil sarita@ibict.br

In the book Memory Practices in the Sciences, Geoffrey C. Bowker discusses the relationship between infrastructures and supporting information used to record knowledge and construct a memory of science (such as manuscripts, printed matter, archives and data bases, among others); and the development of knowledge and information itself. The author not only considers the technical aspects of this debate, but also their socio- political dimensions.

Executive director of the Centre for Science, Tech- nology and Society of the University of Santa Clara, in Silicon Valley, California, Bowker was the first Regis and Dianne McKenna Professor in Science, Technology and Society. Previously, Bowker worked in the Centre de Sociologie de l’Innovation, Paris and the Graduate School of Library and Information Science in Urbana- Champaign, Illinois. Furthermore, he was a Professor at the Department of Communication at the University of California, San Diego.

With Memory Practices in the Sciences, he gives con- tinuity to his academic project initiated in previous works, especially Sorting Things Out: Classification and Its Consequences (co-authored with Susan Leigh Star, MIT Press, 2000) and Science on the Run: Information Manage- ment and Industrial Science at Schlumberger, 1920-1940 (MIT Press, 1994).
His focus of interest again falls on the field of classi- fication and standardisation and its role within the infra- structure of information, particularly in the development of that which is called scientific “cyberinfrastructure” and its social and organisational features. Bowker is especially concerned about the relationships between these subjects and collaborative scientific work, data sharing and the use of computer science in biodiversity studies. The author defines the central question in this field as being how scientists analyse some sciences that contribute to the subject of biodiversity; how they communicate with each other and among themselves in matters such policy formulation - and in particular how data structures and current practices affect this communication.

Through the use of vast multidisciplinary literature, Bowker considers, in Memory Practices in the Sciences, the examination of two great issues: (1) How do scientists configure their own past, either as individuals, “land creatures”, or as pertaining to a disciplinary line? (2) How do scientists configure the past of their objects - the land, climate and the process of extinction?
The author points out the concerns and inten- tions of science, as a social institution, to create a perfect memory of the past; developing technologies that permit the recording of the pasts’ traces, which, conversely, would be forgotten. He highlights that, by means of these traces, they can be understood better as part of a temporal process and larger space.

On the other hand, he argues that the traces we leave - or the records that we execute - do not neces- sarily correspond to that which we were or to the facts that happened. In truth he deals with a tacit negotiation between us and our future readers, listeners or those who want to come to judge or evaluate us. In the case of texts and other scientific records, many times they tell the history of an ideal past, essentially being concerned that the protocols have been adequately followed.
The act of recording (as for example writing a sci- entific article) occurs in the scope of a set of collective practices - technical, educational, social - articulated furthermore in a tenuous way that the author defines as memory practices. These range from being totally unaware to being super-aware. They confer utility to our past in the present, in a way to better drive the future. For the author, the interesting thing is to characterise and understand how a set of memory practices are ar- ticulated in memory regimes, around relatively constant technologies and practices.

The archive is the unit of space in which memory practices are extended, while memory epochs are their units of time. They are units that make communication and information sharing possible. There is also the con- cern with interoperability, on the one hand, and with the loss of data that is frequently caused with the changes in information technologies, on the other. It is emphasised that information infrastructure evolution - especially in information technologies - affects the ways in which we deal with our past, our ability to recover, reconstruct, and forget it. In reality, we think of the past with instruments of our own time and from our social matrices, as well as technologies that we employ in the present, therefore projecting - including onto nature - our ways of thinking and organising them.

The emergence of each memory epoch occurs in association to the development of new recording me- dia, seeing that the limits between these epochs - oral, manuscripts, mechanical, electronic, digital... - are not very clear or evident. The matter of characterising and analysing the circulation of memories occurs through multiple media which are developed by us.
The transition to new memory practice regimes, as well as the increasing valuation of the record on the side of organisations – was not frequent or trivial throughout history, it involved profound transformations in regard to ways of thinking, on both an individual and social plain.

The memory would then be operated by means of a variety of “technical devices”, amongst which the author highlights classification and standardisation. Clas- sification is a resource that allows us to forget something that we will need to remember afterwards. Standards and protocols are formal procedures and/or techniques that result from negotiations and commitments, being essential for communication and sharing between those who are involved in a network.

Without professing a linear chronological narra- tive, the author conveys this discussion from a historical perspective, analysing the ways of networking support material and technological information and the nature of producing knowledge, throughout the last 200 years. Its starting point is the Industrial Revolution in England, arguing that it is was here that new archiving methods were developed - and the new underlying scientific memory practices.

Industrialisation at the beginning of the nineteenth century, in Europe, is considered an important moment in the increase of historical awareness. In a similar way, the period of 1870-1914 is seen as being particularly significant in the recent history of memory practices, with the intensification of concern about standardisa- tion and classification of information, creating space in the twentieth century for the planning of databases as a central, cultural form.

From there, the author constructs his argument on the basis of scientific analysis of the specific methods of reconfiguring of scientific knowledge at three histori- cal points and in three scientific fields, which he calls “memoryepochs”:thegeologyofthenineteenthcentury, mid-twentieth-century cybernetics and today’s sciences of biodiversity. The current epoch is baptised by the author as “the epoch of potential memory”.

He begins by drawing a parallel between the trans- formations in nineteenth century geological science and the change in perspectives of time and memory since the Industrial Revolution, taking the form of a “second nature” completely “archivable” - monetised time, mechanised time and standardised time, essential to the globalisation process that would deepen in the following century.

Next he comments on the cybernetics of the post World War II era, whose emphasis falls into the func- tional similarities between mind and machine, human and non-human systems. Cybernetics thus began as a metascience to which several disciplines would be subordinated, requiring the development of a common language, however also bringing indignity by scientific memory.

He then covers memory practices in the emergent science of biodiversity. This requires the interaction and cooperation of scientific fields and many diverse geo- graphic regions, generating an enormous mass of data and information, simultaneously making the development of unified databases and the compatibility between many classification systems difficult.
Digital archiving reveals the emergence of a new regimen of technologies that intend to conserve - but also to mold - the experience. This new regimen now permits a new flexibility, a new texture, a new mobility of global to local and vice versa. It furthermore makes possible the aggregation of data involving much more complex and diversified operations - such newness is not a quantitative capacity; already having been developed in the nineteenth century, but principally in its curves and folds, its distinct crossings and intersections.

This empirical-analytical statement serves as a con- ducting wire for the discussion about how memory - and our own conscience - are configured very differently in different infrastructures and information technologies. This material is expressed in a metaphoric way. New mediations, new recording media, new memory pros- theses form a new identity already imbricated in these new media. This same information infrastructure is also used in order to speak of ourselves.

The author seeks to demonstrate the overlapping, the traffic and synchrony between the world (and time) social and the natural (and time) world. He argues that information technology, as much as metaphor and materiality, permits the creation of a second middling nature of indexation and archiving, thus eternalising the present and allowing a better understanding of the past. He thus considers that the database and archive - as material substratums and symbolic artefacts - are central elements in the development of longue durée knowledge in the West in recent centuries.

For Bowker, then, the act of remembering doesn’t lessen recovery of the past. However it is inserted in de- veloping reading concerning current methods. Memory - or the act of remembering - is an instrument of sociali- sation and relation with the world, a way to mold and to act regarding the present. There are traces of the past in every part and the possibility of accessing them is, by itself, heartwarming.

The exercise of memory is, according to the author, transcendent and inherent, political and personal, allow- ing a better understanding of our respect and regarding our insertion in the living and inanimate world, in all its complexity. In a similar way, the archive is not simply a stock of useful facts; it is inserted in our sets of actions that contribute in molding the present.
It is a work of great contemporaneity, proving that the fields of information studies and social science studies have much to contribute to each other’s mutual enrichment, opening a scenario for new transdisciplinary practices.


REVIEW THREE

K8's review Jan 07, 09

bookshelves: academic, lis

I just finished reading Geoffrey Bowker's Memory Practices in the Sciences for the book club over at Reading Information Sciences. I gave myself a break and allowed myself to read it at a leisurely pace, a luxury I rarely have anymore.

At the heart of the book, is Bowker's examination of the ways in which "acts of committing to record" are socially imbued practices affecting our conception of the past in ways that. As a consequence, affect our present and future. Bowker begins with the effects of industrialization on memory practices in geology.

For me, what stands out is: The ways accounting practices used during the Industrial Revolution were transferred onto Geology (and geologic ways of deciphering the age of the earth) by Lyell and others. Lyell argued that the earth itself isn't "irregular," rather that it is just a bad archivist. What Lyell constructed was a spatio-temporal systemization that moved away from seeing the history of the earth as a series as catastrophic events. Methods of organizing work and the workplace were used to explain how the earth ages at regular intervals, rather than by events such as earthquakes and floods. He moved away from creationist theories of the earth's age. Bowker discusses how this (and other geologic theories of the time) privileged Western Europe and aided imperial interests and burgeoning globalization.

Bowker follows this approach to memory practices by looking at cybernetics, and its focus on semantic memory - "memory as pure pattern" (pg 76). Bowker sees cybernetics as working from universalism. Theories of cybernetics during the mid-twentieth century posited that cybernetics removes the need for memory as it worked to create a system that could hold all information equally, like a universal discipline. To work, cybernetics must be general. And through cybernetic systems, history becomes just a part of a classification scheme, a way of patterning. And as a result, Bowker argues, memory needs to be destroyed to create unity .

Bowker writes: "First, past disciplines are destroyed: they need to be created anew from first principles Second, an individual experimenter must destroy his or her knowledge of previous experiments. Third, one result of this double destruction will be the discovery by cybernetics that memory itself is epiphenomenal (pg 101).

As Bowker later notes, though, all data is always collected within a context. In cybernetic systems, context is eliminated, eliminating important information about how and why data has been collected. We lose the reason(s) behind collection methodologies which are almost always discipline specific.

Early in chapter 3, "Databasing the World: Biodiversity and the 2000s," Bowker writes "The miracle of memory in our time is that memory practices are materially rampant, invasive, implicated in the core of our being and of our understanding of the world - and yet we experience them and discourse about them in terms of their ideal ramifications on some hypostasized entity created to void materiality from the equation: the individual, the nation-state, the people, and so forth" (pg 109).

What Bowker broaches, in this chapter, is the way we as humans often take ourselves outside of nature when we create our ways of organizing and controlling nature. In addition, he examines the ways in which western peoples' have created systems that place other peoples (as a way of colonialization) classified as 'nature,' while placing their own practices as 'culture' and outside of these modes of classification.

As the chapter progresses, Bowker leads up to a discussion of how we can read databases both materially and discursively. Drawing from Derrida's discussions of how technologies can create new kinds of 'traces.' From these premises, Bowker argues that to 'read' databases, we need to look at what isn't categorized or classified. What is classified is considered important, politically, economically, ideologically, etc. By examining what is left out of these systems, we can start to form the context of systems. He notes that "If certain kinds of entities are being excluded from entering into the database we are creating, and if those entities share the feature that they are singular in space and time, then we are producing a set of models of the world that - despite its avowed historicity - is constraining us generally to converge on descriptions of the world in terms of repeatable entities: not because the world is so but because this is the nature of our manipulable data structures" (pg 146). What this means is that those entities that are named are studied, and once studied, are considered relevant subjects for future study. The unnamed become unimportant, and the systems created around those things we name are not created to support a place for the unnamed.

In biodiversity work, researchers from multiple disciplines using multiple discipline-specific methods and contexts deposit data into common databases. In addition, the data in the databases themselves need to be accessible when databases are updated technologically. As a result, the use of metadata becomes central. Metadata needs to be flexible enough to be accessible both today and in the future. And, it needs to be able to provide as much context as possible. Developing metadata standards that can encompass all possible futures while still serving our present (and the present and future for everyone) is still far off into the future.