RPS' lazy KO blog--occasional musings about the state of knowledge organization

What has happened to the idea of a paragraph? That clever little invention that parses narrative into semantically related clusters, giving breathlessness to the expression of a part of an idea, and yet, by the pause it introduces at its end, allows the mind breathing room while reading a text--the paragraph is in dire peril my friends. I have grown weary of marking student papers "no 1-sentence paragraphs!!!" and yet now as editor of a so-called scientific journal I find myself deluged with these devices too. Okay, so many of them are in manuscripts that have evolved from doctoral dissertations. Still, someone's dissertation advisor should have said "no 1-sentence paragraphs!!!" I wrack my brain trying to understand how this wonderful device, taught to most of us in third grade (or maybe even earlier) could so easily depart from academe. I suppose a lot of it has to do with word-processors, which make whatever drivel one manages to generate after staring at the monitor for hours look like elegant printed text. Rather than actually expressing an idea, we find ourselves instead filling justified space between paragraph marks. Oh my ... Well, look here folks; a paragraph should be several sentences long. It should begin with a topic sentence, usually the first, which is a sort of exposition, or thesis statement. It should be followed by all of the evidence about that topic (properly supported with references, of course). And then, in good sonata form (see Music Appreciation 101) there should be development, in which you (the author, remember?) add value to the evidence by providing your own synthesis about what it means. And then a paragraph should conclude definitively.

Which reminds me, whatever happened to the literature review (stay tuned) that isn't just a litany of "he-said, she-wrote"?

Greetings from rainy Syracuse. This is a small group (around 20 people at any given moment), but the program and the business meeting have both been fascinating. The 10 papers are all really pithy. Based around the concept of North American pioneers, the programme is here: https://netfiles.uiuc.edu/klabarre/shared/iskona/nasko2009.html for now. As part of my paper developed using author co-citation analysis of North American KO authors I ran up a quick co-citation analysis from the 10 papers: there were two clusters--Smiraglia, Miksa, and Shera in one; Hjorland, Mai, Tennis, Olson, and Bates in the other--the conceptual basis of the clusters seems to be bibliographic classification and fundamentals of KO. The link between the clusters stretches from Shera to Hjorland (which I thought was fascinating). The cluster on facting (La Barre, Cochrane, Richmond, Ranganathan) dropped off while I was trying to make the map readable; when I have more time I'll run it again. Anyway, stay tuned ....

not dendogram ... "Dendro" from the Greek "dendron" for tree ....

although I was tempted to say: "Dendro" named for its inventor Sir Richard Dendro the famous irish physicist. Let's see if this gets into Wikipedia.

My paper for the ethics symposium (see my last post) was called "Bibliocentrism, Cultural Warrant, and the Ethics of Resource Description." The point of the paper was that the pervasiveness of bibliocentrism constitutes an act of harm against the user-community by preventing Wilsonian exploitation. As usual, I don't think I said that in my presentation (but I did say that in the paper). My method was case study--I took 3 "biographies" and 4 "nonbooks" and tried to demonstrate the uselessness of title-page transcription for facilitating exploitation. The connection being, of course, that bibliocentrism leads to standards that treat everything like a book (because books are good and nonbooks are nongood) and does even that badly.

I think listeners probably got the idea that I think title-page transcription should be replaced by digital imaging. That's correct. I do think that. I used Morgan's FDR as an example, to demonstrate how one could link from the table of contents--say concerning the Newport Incident--to the source list and then directly to the Newport archives where the sources are held. But there is much more, of course.

The ethical issue is that bibliocentrism, like racism or sexism or most 'isms," is endemic, and is perpetuated by the participation, tacit or otherwise, of everyone in the resource description community. An ethical response requires first an admission of complicity and then a commitment to change.

I ended the presentation (but not the paper) with an image of Otlet's "grinder," simply to demonstrate one possibility for an alternative to a bibliocentric catalog. (Documents go into the grinder to be disassembled into facts, which come out newly synthesized as raw knowledge, classified using UDC. See my new favorite book: European modernism and the information society, ed. by Boyd Rayward, for both the image--the grinder is on the cover--and more explanation.)

But, there is still more. I've been working with Jeff Gabel on his citation-chasing project, specifically now on the second phase which uses MDS to analyze co-assignment of LCSH to citation-chased monographs. (I'll post something specific about this soon, so bear with me.) If this technique were to be employed, it would require systems that have all citations in digital form, and I don't mean in separate citation indexes. So one very useful thing a non-bibliocentric catalog could do would be link to the digitized citations in everything.

I'm just home from Milwaukee, where for the past two days a very intense conference on ethics and the organization of information took place. The abstracts and program can be found following the link below. The papers will be published in a theme issue of CCQ before too long; for now the abstracts are on the website but I think the slides will be posted soon as well.

On my way home, sitting at the airport in Milwaukee (a very contemplative spot), I just kept thinking the whole issue is like a huge morass, like hundreds of nested issuespaces, each with its own structural critical points.
Example--Janet Swan Hill's metaphor of copy cataloging being like a chain letter. If everybody follows the directions everybody gets lots of goodies. But even if all but one follow the directions, somebody will get metaphorically screwed. So when one library decides not to input original cataloging it potentially brings down the whole system. Causing harm--that's at the core of ethical thought. How is this a nested morass? Standards are too hard to apply, standards aren't empirically-based, standards aren't really user-centric, librarians don't apply the standards. But those librarians are the ones who created the standards. The educational apparatus isn't doing its job; the professional apparatus isn't doing its job; the bibliographical apparatus is weakened; the users suffer. Yet everyone is working very hard and doing their jobs the best they can. Where to begin? Pick a spot, and figure out how to do no harm.

My own topic (bibliocentrism) in another entry soon.Information Organization Ethics Conference Fortunately, I'll be having more contemplative time at General Mitchell airport soon. A very contemplative place.

I absolutely believe in the double-blind peer-review system for advancing scholarship. I have experienced the system from every which way possible, and I have absolute faith as a consumer of research (not a reader, but one who uses scholarship to advance my own scholarship) that peer review has led me to valid data. I am writing a paper now, for instance, in which I am relying on a paper by Maria Lopez-Huertas; I know I can count on the validity of her data to inform my own data analysis.

Still and all, the system has its quirks. I was astonished and dismayed to discover that JASIST was no longer using double-blind review. I discovered this, to my dismay, when I was sent a paper for review that turned out to be a paper by one of my own students that I had given a less than wonderful grade. Of course, I'd have recognized the paper anyway; but I was appalled to receive it unblinded, as it were, for review.

Most of the journals I read for maintain double-blind review and I appreciate it. Of course, there is often a moment when one thinks one knows who the author is, but the polite thing to do is put that thought out of your head and proceed as though you didn't know (and who knows, you might not).

Reviewing in the knowledge organization domain also has its own domain-centric characteristics. For one thing, we are a small domain with a lot of ongoing work. Every year there are regional conferences and every other year there is an international conference, so there is an almost constant demand for 60 or so referees to be reading. I have two really terrific referees, both of whom return papers to me at once (usually overnight, but occasionally within a couple of hours). I figured out they both are simply reading them as they arrive in the email and therefore getting them out of the way. I have adopted that practice as well, and I'm much relieved not to have an inbox full of papers for review. I recommend this approach highly.

I'm always irritated when referees turn me down; I figure, we're all in this thing together and we all have to play our parts, whether the dog is sick or not. But, it happens.

Papers for Knowledge Organization are sent to three referees. Most reply within a month, although in rare cases I have to chase after a reader. In some cases I never hear from the person. I enjoy getting diverse reports (a hates it, b loves it, and c thinks it needs work) because usually it gives me a fair amount of leeway for advising the authors. Sometimes referees get too wrapped up in grammar and punctuation. I figure that is the editor's job--a referee should comment on the originality of the research, its appropriateness for KO, the rigor of the methodology and accuracy of the results, and applicability of the conclusions. Referees also ought to check the references, not necessarily for formatting, but for the presence or absence of material that ought to be cited. After all, this is how a domain acquires cohesion. This is the gatekeeping function that constantly checks the intension of the domain.

About fifteen yeras ago I was asked to consult on a project called the Union Catalog of Motion Picture Music. The project has not advanced, although the research it was invented to serve has done so nicely. I'm once again visiting with these folks at a symposium at USC concerning musicological film studies.

There seem to be two essential problems for this domain--the musicological study of film music. First, that sources are missing, lost, dispersed, or non-existent. The second is that the sources are not of a traditional sort. This is to say, there probably are no definitive full scores, and of what does exist, much is in private or industry hands or has been lost. So here are two pdf's--the symposium program, and a pdf of my PowerPoint slides (with references added at the end).

As editor of the journal Knowledge Organization I have been observing the citations in various papers by different authors. My casual observation that there was little conformity among authors has led to a few small studies of literature in knowledge organization. It is particularly simple to analyze the proceedings of individual conferences to discover whether the papers illustrate any common core. Last summer's North American Symposium was discussed in v. 34 no. 2; this summer's 10th International Conference is summarized in v. 35 no. 4. For the latter I produced an Excel spreadsheet, which is somewhat messy, but you can download it from here should you want to play with it.

The relatively new domain of Music Information Retrieval or MIR is a rapidly evolving, technology-driven recent entrant on the information retrieval scene. Generated by information scientists, computer scientists, engineers, mathematicians, and musicologists, among others, the domain has contributed new systems for automatic storage and retrieval of music. Mapping the domain is itself a fascinating business. Recently I asked "Music Information Retrieval: An Example of Bates' Substrate?" in a paper for the Canadian Association for Information Science/L'Association canadienne des sciences de l'information. This is the abstract:

Bates suggested that the intrinsic unity of information science lies in 'substrate'-the properties of information and its transmission. Music Information Retrieval (MIR), and ISMIR annual conferences offer a rich panoply of intellectual and cultural diversity. We map the evolution of MIR using conference papers from 2000 through 2005. Results indicate tight thematic coherence in the domain around the problems of information retrieval and classification, and the locus of most research within computer science departments.

The paper is available here: http://www.cais-acsi.ca/search.asp?year=2006.

Author co-citation analysis was also revealing: indicat[ing] tight thematic coherence in the domain around the problems of information retrieval and classification, and the locus of most research within computer science departments. Citation practice indicates the habits of a hard science. Author co-citation within the domain is abundant, J. Stephen Downie is clearly the founding focal point, but the domain is very focused, reinforcing the notion of a tightly-packed, emerging and continuously successful domain. ACA data from outside the domain provides an interesting comparison; watch for another paper soon.

Inspired by encountering quotations from Two Kinds of Power in conference papers last summer I undertook an analysis of the domain defined by those who cite Wilsons famous book. The paper is to be presented at the 2007 conference of the Canadian Association for Information Science/L'Association canadienne des sciences de l'information: Among Patrick Wilson's most influential books was Two Kinds of Power, which has influenced scholars in information science, and particularly in knowledge organization. Tools of domain analysis are used to analyze the corpus of literature that cites Two kinds of power. Aboutness and relevance are demonstrated keys to this specialization.

The Proceedings are here: http://www.cais-acsi.ca/2007proceedings.htm. I am really quite fascinated by the concept that author co-citation analysis gives us a picture of symbolic interaction. That is, that what we see is how the scholarly community perceives intellectual connections among the co-cited authors. As I mentioned only briefly in the paper, there seem to be clear social networks in the map that focus on the lineage of dissertation advising. For my presentation I added a final slide using this quotation from Two Kinds of Power (p. 132): "Let us imagine a Supreme Bibliographical Council, whose task it was to evaluate the bibliographical situations ...." I decided that's what we're looking at here. Marcia Bates is the "chief justice" and there are two parties, one in IR represented by Belkin and Saracevic, and another in KO dominated by Hjorland, with Howard White as the swing vote. Well, it's a metaphor ....

Fascinating to see Åström's paper in JASIST 58(7): 947-57, in which he finds informetrics and ISR stable but that user-oriented and experimental IR research have merged into one field--ISR. This is comparable I think, to my finding that "aboutness" was a historical node but has given way to IR and KO. Interesting ....

In the meantime I prepared a paper on superworks for a volume edited by Arlene Taylor <here>. This was fun, because often when I talk about cultural forces as catalyst for instantiation peoples' eyes glaze over, but this time I was able to use Brokeback Mountain as an example. It had been fun to see the movie in the US one week, and a week later find a novelization in English with the movie actors on the cover in a bookstore in Amsterdam.

Constellations of works exist with abundance in the bibliographic universe. While this is good news for library users-cultural forces drive the marketplace to see to it that a wide variety of useful instantiations evolves-it presents a challenge for information retrieval. A simple citation for a work might be the anchoring node for a large family of related works. The future of sophisticated information retrieval depends on the development of integrated repositories that allow informed selection among the plethora of entities that share intellectual content. Achieving this goal will bring us much closer to Wilson's notion of exploitative control of humankind's store of recorded knowledge.

I recently published a larger meta-analysis of the concept of instantiation (A meta-analysis of instantiation as a phenomenon of information objects. Culture del testo e del documento 9, no.25, gennaio-aprile 2008, pp. 5-25). Here is the abstract:

"Instantiation" is the phenomenon observed among information objects of all types, in which multiple iterations of the information content exist and must be collocated and disambiguated in a retrieval system. The phenomenon has been observed among bibliographic works, cultural heritage artifacts, archival documents, scientific models, and ontological constructs. Studies have demonstrated some consistent theoretical parameters for the concept of instantiation, such as the importance of canonicity as a catalyst for instantiation, positive correlation of age of progenitor with large instantiation sets, and positive correlation of age of progenitor with complexity of instantiation sets. In the present paper all relevant terms are defined, an epistemological analysis of the concept of instantiation is presented in summary form, and a meta-analysis of the phenomenon of instantiation is performed using empirical evidence from several studies. The result demonstrates theoretical consistency across studies, suggesting the importance of the phenomenon for the development of the semantic Web, as well as pan- and inter-institutional digital libraries incorporating representations of both documentary and artifactual information resources.

At the moment I am interested in understanding the concept of instantiation as a form of metonymy.

I attended the workshop Can You See What I Know? (http://cyswik.blogspot.com/) presented by the Virtual Knowledge Studio in Amsterdam (http://www.virtualknowledgestudio.nl/). If you have time to watch the videos you can see me eating (tuna-salad sandwich) in the first one, and in the second one I actually get to talk about disturbance as a catalyst for knowledge acquisition. At least, that's what I meant to say.

CYSWIK was a remarkable two-day event, bringing together artists and scientists and humanists for discussion and learning. It was a socially difficult activity because of the different vocabularies and modes of thought across the domains. But I think we all learned an immense amount, about ourselves especially.

A brainstorm with Charles van der Heuvel of the VKS (http://www.virtualknowledgestudio.nl/staff/charles-van-den-heuvel/) will lead to a collaboration on what we are so far calling an "idea collider." I hope this will incorporate the ontological footprint technique. Stay tuned for updates.

In the exercise of developing the CIDOC-CRM it became apparent that using the ontology to map information objects would reveal certain patterns of entities, properties, and relationships.
Furthermore, these patterns, when analyzed, reveal essential footprints of information objects. That is, like a genome, a CRM mapping records the essential informative properties of mapped
objects. The region for research here is pure theory. What categories can be observed among mapped information objects? When is a sailor’s deck-log like a terracotta hut urn? I have constituted a research team and with a very small grant from Long Island University we have begun developing techniques for mapping, and a calculus for data-mining the maps in order to generate clusters (or classes) of information objects. We had one poster at the ISKO conference this summer ("Classifying Information Objects: An Exploratory Ontological Excursion," by Sergey Zherebchevsky, Nicolette Ceo, Michiko Tanaka, David Jank, Richard P. Smiraglia, and Stephen Stead. Poster presented at 10th International ISKO Conference, Montréal, 5-8 August 2008). The poster can be seen here (or have a look at these pdfs from ISKO and ASIST).

Edmund Husserl's phenomenology is just one of the 20th century's fascinating schools of philosophical thought that is directly relevant to notions of knowledge and information. I have been experimenting with the differentiation of otherwise like entities by attempting to identify their perceptual differences--Husserl calls this noesis, or the act of perception through one's own ego. Trivial examples are most entertaining so here are two. This sign from a hotel in Amsterdam puzzled me for years:

The text says "Wat to doen bij brand," which as you can see means "what to do in case of fire." (The photo is fuzzy, which is a shame, because it means I will have to fly back to Amsterdam to take a better picture!) The first time I encountered this I was rather jet-lagged and thought "how odd, instead of running apparently you are supposed to scream." What did I know of Dutch culture? But the more times I pointed this out to people in the hotel the more times they said to me "looks like he's dancing." So there you have it--the picture shows a person by a fire. I see someone screaming, many others see a person dancing. Those are ego-acts--noesis--self-experiential interpretations. It is one reason classification can be so difficult, because the same thing can mean different things to different people. Here is a set of pictures from the intersection of Frauentorgrabe and Kartäusergaße in Nürnberg; this is where you turn to approach the Germanisches Nationalmuseum:



Clearly, in Germany in order to cross the street one must stand atop a bicycle. Note that if the light is red one is compelled to balance there at rest until it turns green. Obviously this is problematic for some citizens--the older gentleman in the last photo has acquired the requisite bicycle, but although the light is green he is hesitant to leap onto the bar to cross the street.

Okay--point made? I always think it is an interesting philosophical exercise to approach a scene as though one were a creature from outer space and ask oneself "what am I seeing here?" Do you see that long line of earthlings on the right in that third picture? They have evolved to a high capacity and even seem to float as though on wheels; when they become excited their eyes shine enough to brighten even the night. Unfortunately, all of them are infested with two-legged parasites. Their civilization must learn to deal with these infestations before we can settle among them.

(Are you curious about how much traffic I stopped taking these pictures? I'm interested to know whether anyone has noticed I've taken hundreds of infrastructure photos around the world recently as part of this study!)

What is the use of this research stream? At present an obvious implication is the explanation of divergence in Web 2.0 applications--when is a tag meaningful and to whom? But there is much more potential here as well. For instance, in my paper for last summer's ISKO conference I developed the idea of noesis as the synthesis of perception. Here is the abstract:

Perception is a crucial element in the viability of any knowledge organization system because it acts as a filter that provides contextual information about phenomena, including potential categorical membership. Perception is moderated culturally, but "social" systems exercise little or no cultural conformity. "Every day classification" is rife throughout human experience; but classification arises as a system of formal constraints that embody cultural assumptions about the categories that are the products of human cognition. Noesis is a perceptual component of Husserl's phenomenological approach to human experience. How we perceive a thing is filtered by our experiential feelings about it. The purpose of this research is to increase understanding of the role of cognition in every day classification by developing a fuller profile of perception. Photographs of mailboxes (a mundane, every-day example) from different locales are compared to demonstrate the noetic process. Tag clouds are analyzed to demonstrate the kinds of perceptual differences that suggest different user perceptions among those contributing tags.

("Noesis: Perception and Every Day Classification." In Arsenault, Clément, and Tennis, Joseph, eds. 2008. Culture and identity in knowledge organization: Proceedings of the 10th International ISKO Conference, Montréal, 5-8 August 2008. Advances in knowledge organization 11. Würzburg: Ergon Verlag, pp. 249-53.)

While trying to illustrate this process I realized that the heretofore supposed origin of perception is not in the information object itself, but rather is in each person who interprets it. So this accords with the phenomenon of instantiation. Instantiation says there are many perceivable iterations of information, and phenomenology says there are many potential noetic acts of perception. What is the chance that any two of these streams will meet in a human mind and form an understandable chain? A million research questions now follow; stay tuned.

Just a joke (sort of). In 1987 when I was teaching at Columbia University, I had two computers in the room, projecting in different directions, and chalk boards. So I could have OCLC in one, a word processor in the other, and make notes to emphasize my lecture on the board. This past week, after 16 years at LIU, I finally had that functionality. Just in time to leave. What comes around goes around.

Sometimes it's like pulling teeth ....

Okay, so I've been working with noesis, which is a matter of perception. The work all boils down to a basic question: how do we organize knowledge conceptually so long as every concept is perceived differently by everybody? The answer is that the whole process is a huge and constant brain-massage, wherein we shove a little to the left and then a little to the right and back and forth and on and on, trying to get everybody to agree to a common set of perceptions. Which, of course, was the idea behind universal KOS.

I thought this cartoon was perfect:

In Spring of 2008 there was a lively discussion on ISKO-L concerning Zins' method for creating the pillars in his worldmap of human knowledge, which is a sort of classification/encyclopedia KOS. Sort of. Here are the comments as I archived them at the time.

Leon Manifesto

Chaim Zins and the Map of Human Knowledge.

See his paper in the 2006 ISKO Proceedings. Then please visit the site using the link I think I've added.

(Sigh, this software is a little odd.)

Worldmap