A very typical sort of Information Science article in a scholarly publication begins something like this:  "What is information?  It is that which reduces uncertainty.  It is that which assists in decision-making?" (1)

Or take this example from page one of an introductory information retrieval text:  "To know what information retrieval is, we must first know what information is.  There is no fully satisfactory definition.…Temporarily, consider the oversimplified characteristics that information is something that (1) is represented by a set of symbols, (2) which are organized or fit into some structure, and (3) can be read and to some extent understood by users of information" (2).

One would be hard pressed to find another field of scientific inquiry in which every researcher gets to define the primary object of study for himself.  Imagine if every chemist had to state to which definition of matter she was going to adhere, or if a biologist was forced to define life at the beginning of every lab report.  This last example leads me to my point:  there was a time when a chemist would have started an explanation of a new experiment or theory with just such a definition.   Information science is currently in such a time.  We have yet to agree on a core set of concepts.  Our facts are in dispute.  Our vocabulary is in a high state of flux.  We are trying to define our object of study.  None of these makes what we do any less of a science.

Science and Technology Studies (STS) practitioners would argue that we are a prime target for their inquiries.  Nearly everything we do happens 'on the edge' of our science--what Jay Labinger calls "frontier science" (3).  First of all, we are an extremely new science—the term information science only supplanted documentation in the early 1960s, and not officially until 1968, with the American Documentation Institute changing its name to the American Society for Information Science (4).   Second, we are hopelessly, and quite possibly forever, enmeshed with our parent field of librarianship.  How closely Norbert Wiener and Vannevar Bush identified themselves with librarians is a matter of pure speculation.  How closely we identify ourselves with them seems to be 20 / 20 hindsight. 

It is in this atmosphere then, that we are trying to stabilize our core set of concepts, evolve criteria by which to judge the quality of research, and even more basically, decide on appropriate topics for research. STS does most of its work on this type of frontier precisely because of this high rate of change:  studying the evolution of knowledge during times of great change is much more interesting than studying it during times of stasis.   Certainly, this edge is also where scientists are most often wrong, but the job of STS is to comment on the process of science--not the validity of individual results. 

STS’s earliest work was by feminist anthropologists casting a critical eye toward the practice of anthropology (5).   Sociologists and anthropologists then turned this technique toward the observation of natural scientists in their natural habitats:  in labs and in the field. Historians of science, philosophers, cultural theorists, and literary critics have also come to embrace some of these methods, searching for science's influence on society at large, and vice versa.  Along the way, controversy has arisen as some scientists have revolted against being treated as lab rats by those with no scientific credentials, and others have ridiculed the whole exercise as “merely incomprehensible language masquerading as knowledge” (6). 

These controversies have become known collectively as the Science Wars, and rather than add to their chronicles, I would like to take a couple of key models from Bruno Latour’s Pandora’s Hope, which exhaustively defends the work of science studies by giving it a thorough grounding in epistemology, phenomenology, and semiotics (7). 

The first model is that of circulating chains of reference.  This should interest us, if not just for its great name, then because much of what we do is managing the last links of all the sciences’ chains--the journal articles, conference proceedings, and monographs that are the end result of field work and laboratory experiments.  We are intimately familiar with these products—so familiar that we frequently mistake them as being in our own domain, when we are only charged with ensuring their safekeeping and ready access.  At the opposite end, the first link of their chain is some measurement of an object in the world.  This measurement is the translation of a fact, through some sort of an instrument, into a word or a number.  As the world slowly moves onto the shelves of our libraries, "there is...a complete rupture at each stage between the 'thing' part of each object and its 'sign' part" (8).   Latour's example makes this chain seem simple to trace:  the junction between a savanna and a forest is observed on a map laying on a table in a restaurant in South America; a grid is then transposed onto the landscape by pounding stakes into the ground at points of the grid (found by measuring the distance from previously mapped trees with a device jokingly referred to as a pedofil, the pedological translation of Topofil Chaix, a surveyor’s tool adapted to forestry), soil samples are taken at points along the grid, deposited into a matrix of cubes (called a pedocomparator) that correspond to their original locations, brought back to the restaurant tabletop for simple ‘tasting’ and color analysis.  Finally conclusions are drawn, quite literally, in a journal article whose main illustration was produced on that same restaurant table where we began.

At each of these stages, one 'thing' stands in for another.  The diagram and text produced in the above example bear little resemblance to the forest / savanna juncture, but by tracing the origin of each reference, one is able to move back toward the field and verify the scientists’ conclusions.  "Like the footnotes used in scholarly works to which the inquisitive or the skeptical 'make reference'…, this armful of specimens will guarantee the text that results from [the] field expedition" (9). 

These inquisitive and skeptical people use our tools, our chains of reference, to trace these citations—backward chaining in librarian-speak.  A simple chain of reference for a library is easy to articulate:  a library decides what sorts of materials to collect based on the wishes of its constituencies; some sort of formal collection policy is established; classification schemes are picked so that topics can be explored via surrogates instead of having to navigate the source materials directly;  a storage scheme is picked and (possibly) mapped onto a physical structure so that individual items can be found once they are identified. 

A more generic information storage and retrieval system might be a little harder to understand in this way, but the basic purpose is the same: to erase the boundaries between descriptions and the entities described.  In an ideal system, we would "never detect the rupture between things and signs…We [would] see only an unbroken series of well-nested elements, each of which plays the role of sign for the previous one and of thing for the succeeding one" (10). 

Each of these elements can be called an actant, a term Latour borrows from semiotics.  An actant is any participant in an endeavor, whether human or nonhuman.  An actant is defined by what it does, in essence, what function it performs within a discipline.  It emerges through trials (experiments, tradition, practical application), and once its fellows (usually human actants in the discipline) deem that it performs adequately (through a regularized process, such as peer review or commodification), it is admitted to an institution.

The chain of reference described above was said to circulate, and while reference materials do not ordinarily circulate, Bruno Latour’s do.  In his circulatory system, actants move along five intertwined loops: mobilization of the world, autonomization, alliances, public representation, and links or knots between these loops. 

Links and knots are perhaps the easiest part of Latour’s circulatory system to understand, even though he fights against defining them too precisely.  These connections between the other four loops exist because they are responsible for constantly feeding the whole system with the content of science—the core concepts, practices, and intellectual products of science live in these knots.  Without these links and knots, there are no other loops, but at the same time, without the other loops, the content of science withers away and dies.  This loop obviously includes our definition-evading information.  It also includes all those theories and tools that we borrow from other disciplines: information behavior (simply ethnographic studies from anthropology), information transfer (communication theory by any other name), systems theory (from whatever circle of hell will claim it), and our complete mis-reading of information theory (from engineering and mathematics).

It is the combination of these actants that created a science out of documentation.  After the amount of scientific information exploded during World War II, engineers, chemists, and documentalists invented systems to handle the ever-increasing flow of scientific papers: selective dissemination of information (SDI), Keyword-in-Context and Keyword-out-of-Context machine indexing (KWIC/KWOC), Index Chemicus, and ERIC are all techniques and systems that have their genesis outside of the library (11). 

Notice that no 'native' library science concepts were included in the above discussion.  This is simply because it is so hard to find them.  If it is starting to become apparent that information science is a strange hybrid field, it must also be apparent that library science never was.  From the outset, librarians might simply have been caretakers and educators for information, rather than students of it.  As the chemists, engineers, mathematicians, and cyberneticists began to look for a home, perhaps library science was simply the easiest domain to conquer.  Or perhaps, as I believe, librarianship looked at this rapidly evolving field and saw its ability to manipulate information in ways that are obviously good, and saw its own reflection.  It certainly adopted the new tools and methods quickly enough.  However, if we now abandon libraries altogether by putting all of our efforts into the market economy, if we become slaves to private research dollars just because that is where we are currently having the most success, and if we convert our buildings to massive public-access computing facilities, we will be in danger of losing not only our best public face (see the next  loop), but also the only collective history to which we can lay claim. 

The second loop, public representation, is frequently denied by scientists. (As if it were possible for science to exist in a vacuum.  If nothing else, science studies has done a very good job of putting that notion to rest.) Public representation is an easy concept to grasp: science has a public face, and this face must be acknowledged and it must be carefully protected so that we can continue our work.  For the past five years information science has been able to take advantage of the emergence of the Internet as a public phenomenon, but this actant is currently taking a beating at the hand of viruses, child pornography, failing dot-coms, and the abuse of privacy.  Perhaps search engines will emerge as the exemplary public face of information science.  Certainly they are the biggest implementation of IR ever, and have done more to bring concepts like recall and relevancy to a wider audience than any bibliographic instruction program ever has.  This notoriety is, of course, a two-edged sword, as poor precision and excessive recall are not concepts we necessarily want to hear people talking about in pubs and on busses.

Alliances form the third loop of our circulatory system.  These alliances are the interactions that the field has outside of its immediate sphere of influence that are required to get work done.  This is something that must be created, Latour argues, because the connections are never self-evident or natural.  Certainly we continue to have strong ties to our communities through our public libraries.  Medicine, engineering, and the natural sciences continue to be strong clients--if only because they would collapse under the weight of their own information without us.

A new look for us, rather shocking for some, is that many practitioners are moving into the private sector.  Perhaps these frequently young and recently trained information professionals are the advanced guard for information science as they seek the more lucrative pastures of the corporate world.  As private organizations increasingly replace the government and large institutional libraries as sponsors of research (12),  they will be more comfortable if this inclination toward information science has been "made to appear, in retrospect, inevitable,” by our advanced scouts.  The goal here can be seen as placing the discipline “in a context sufficiently large and secure to enable it to exist and endure” (13).  The danger, in my opinion, is in casting the net of context so widely that it loses its shape and focus.  If information is everywhere and constantly in use and it is all within our purview, then what is out of our scope?  At that point, what would separate us from any other discipline? 

The fourth loop, autonomization, is the process by which “a discipline, a profession, a clique, or an ‘invisible college’ becomes independent and forms its own criteria of evaluation and relevance” (14). This loop also involves the formation of associations and schools, and it is this loop that is currently causing us so much pain regarding what we call ourselves, and how we train ourselves.  But just as the Science Wars have barely registered with scientists, and almost not at all with the general public, the L-word Wars are but a minor blip.

Let us be realistic.  Library science, as I hope I have shown, was always a misnomer.  Information science is not much better, but at least it is better than documentation.  It gives librarians someplace to hang their hats in the academy along with others that seek to study what information is, how it acts, and where it accumulates.  Librarians, along with a host of other people that do not self-identify as information scientists, are responsible for preserving these accumulations of our human memory in all of their forms, are responsible for ensuring that scholars, ordinary folk, workers, and future generations all have access to the information they require, or about which they are simply curious.  Whether we refer to information bound to a page by Gutenberg, or freed into the ether by Claude Shannon, we are responsible for it, and must do our best to be good stewards.

This leads us to our final loop: mobilization of the world.  This refers to the ability of science to move pieces of the natural world around with us.  The actants in this group are the instruments and equipment that allow us to measure, detect, sample, and describe.  For information science, this obviously includes the tools we use to transfer documents and descriptions of documents with ease--in essence, the tools we use to make information move. 

At the same time, the motion produced by these actants seems to be forming yet another definition of information.  What moves is not so much measurements, detections, samples, and descriptions.  What we move is information. 

As information moves through the world, we can, in turn, measure it and its motion.  It tends to aggregate in great big piles: in libraries, on the Internet, in large organizations.  These aggregations, in turn, tend toward autonomy--with their own rules of evaluation and relevance.  Along the way, our information in motion forms alliances with other actants to enhance its ability to move, to increase its power in the world, and thereby increase its chances of survival.  The faster it moves, the more power and influence that it has, the more likely it is to be noticed by the public. 

Now that we have watched information in motion, circulating back through the system, we can see that information itself forms the links and knots at the center of Latour's model.  Now it becomes obvious that information is the core concept that binds our science together, and the urge to define it once and for all, to confine it to one standard model, is not the end goal of our endeavor, nor is it necessarily our starting point.  The imaginary biologist with whom we began this paper is not forced to define life at the beginning of every lab report.  To be sure, there are biologists that debate the issue, but it does not prevent them from doing work.

1. Sylvia G. Faibisoff and Donald P. Ely, "Information and Information Needs," Information Reports and Bibliographies 5, no.5 (1976): 2.

2. Charles T. Meadow, Bert R. Boyce, and Donald H. Kraft, Text Information Retrieval Systems, 2d ed.  (San Diego: Academic Press, 2000), 1. 

3. Jay A. Labinger, "The science wars and the future of the American academic profession," Daedalus 126, no.4 (1997): 214. 

4. Dorothy B. Lilley and Ronal W. Trice, A History of Information Science 1945-1985, (San Diego: Academic Press, 1989): 4.

5. Sara Franklin, "Science as Culture, Culture as Science," Annual Review of Anthropology 24 (1995): 169.

6. Joan H. Fujimura, "Authorizing Knowledge in Science and Anthropology," American Anthropologist 100, no.2 (1998): 348 

7. It might be this thoroughness that gives so many scientists pause when reading works such as this—it is not the most accessible book in the world.  Latour himself posits that “good scientists enlist in the science wars only in their spare time or when they are retired or have run out of grant money,” and “scientists spend only a fraction of their time purifying their sciences and, frankly, do not give a damn about the philosophers of science coming to their rescue.” Bruno Latour, Pandora's Hope: Essays on the Reality of Science Studies, (Cambridge, MA: Harvard University Press, 1999): 19 

8. Ibid., 60.

9. Ibid., 34.

10. Ibid., 56.

11. Lilley, ch. 2.

12. I include large institutional libraries as centers of research because they so often implement projects of large scope that frequently appear in the literature.  These libraries also form alliances with technical experts who may, but frequently do not, consider themselves information scientists.

13. Latour, 104.

14. Latour, 102.

Faibisoff, Sylvia G. and Donald P. Ely.   "Information and Information Needs." Information Reports and Bibliographies 5, no. 5 (1976)

Franklin, Sara. "Science as Culture, Culture as Science." Annual Review of Anthropology 24 (1995): 163-184.

Fujimura, Joan H.  "Authorizing Knowledge in Science and Anthropology." American Anthropologist 100, no. 2 (1998):  347-360.

Labinger, Jay A.  "The Science Wars and the Future of the American Academic Profession." Daedalus 126  no. 4 (1997): 220-221.

Latour, Bruno.  Pandora's Hope: Essays on the Reality of Science Studies.   Cambridge, MA: Harvard University Press, 1999.

Lilley, Dorothy B. and Ronal W. Trice.  A History of Information Science 1945-1985.  San Diego: Academic Press, 1989.

Meadow, Charles T., Bert R. Boyce, and Donald H. Kraft.   Text Information Retrieval Systems, 2d ed.  San Diego: Academic Press, 2000.