Tuesday, January 29, 2008
Smith, "Cataloging and You"
S
mith, Tiffany L. "Cataloging and You: Measuring the Efficacy of a Folksonomy for Subject Analysis." Ed. Joan Lussky. Proceedings 18th Workshop of the American Society for Information Science and Technology Special Interest Group in Classification Research, Milwaukee, Wisconsin. 2007.
Smith establishes a basis for comparison between the user-generated folksonomies developed in association with Library Thing and the Library of Congress Subject Headings for the same works. Her central research question attempts to reconcile each of the systems with matters of "efficacy and accuracy." In these terms, both folksonomies and the LCSH system have their limitations: folksonomies area hampered by variability (no shared vocabulary is imposed where folksonomies flourish); on the other hand, the LCSH is challenged by "currency, exclusions, and latencies" (para. 6). Smith explains each of these limitations in a fair amount of detail (paras. 7-10). She notes that the LCSH system is slow to adapt (might its inertia be its purported strength?), and yet, the flexible vocabularies we find in folksonomic classification tends to introduce redundancy that might mischaracterize and, therefore, mislead.
Smith also accounts for the problem of inflexible categorization schemes and latency. Controlled vocabularies cannot adapt to that which has never been done before. Another limitation for the LCSH is what she calls "pre-coordinate indexing" (a synonym, I assume, for the preformed taxonomy):
Pre-coordinate indexing forces the cataloger to prognosticate in relation to what future users will find of value in the information entity. There will necessarily be some aspects of every text that the cataloger does not include. The problem, of course, is that these areas constitute latencies of the book's subject that may compromise retrieval of information. This is further exacerbated by the issue of catalogers' quotas and a contributing issue: we don't get to read the entirety of most of the books that we catalog. (para. 11)
The point about not reading and cataloging or partially reading and cataloging introduces an intriguing twist here: What sort of knowledge is involved in the act of classification in either system? How greatly does this knowledge differ? And is it the varying thickness of this knowledge (re: thin slicing) what unsettles skeptics of folksonomic classification systems (as popular, participatory method)?
Within this long-ish quotation, I am also interested in the notion of a system that tends to stagnate because it cannot anticipate the scholars of the future. Derrida gets at this in Archive Fever, and it would be interesting to look at this tension against Carolyn Steedman's treatment (rebuttal, of sorts) of AF in Dust. How does Dust deal with classification or position the "breath it in" archivist as one whose indexical acts carry forward (draped in ethics, anticipation, and so on)?
Back to the article: Smith identifies her comparative approach as "exploratory" and "crude," and although I have a different interest in tagging practices than hers (efficacy and accuracy), I regard this as a solid overview, one well-grounded in a promising lit review (see below) that makes sense of the relationship between taxonomy and folksonomy relative to a smart Web 2.0 application in Library Thing. Smith's methods are visible on a different scale in the second half of the essay, where she works through the comparisons of five books according to how they are labeled in each of the systems.
Smith's lit review is one of the strong points of this piece. Here are a few items from the works cited that stand out to me, and that I will track down when I return to questions of tagging practices (how best to describe them, differentiate them, teach others about them, etc.) in revisions of Chapter Three:
Bowker, Geoffrey C., and Susan Leigh Star. Sorting Things Out: Classification and Its Consequences. Cambridge, MA: MIT Press, 1999.
Guy, Marieke, and Emma Tonkin. "Folksonomies: Tidying Up Tags?" D-Lib Magazine 1 (2005), 24 Apr. 2007 < http://dlib.org/dlib/january06/guy/01guy.html>.
Hammond, Tony. "Social Bookmarking Tools (I): a General Review." D-Lib Magazine 11 (2005). 24 Jan. 2007
O'Connor, Brian C. Explorations in Indexing and Abstracting: Pointing, Virtue, and Power. Englewood, Colorado: Libraries Unlimited, 1996.
Svenonius, Elaine. The Intellectual Foundation of Information Organization. Cambridge, MA: MIT Press, 2001.
Postied by at Wednesday, June 20, 2007
Flower and Hayes, "Problem-Solving Strategies and the Writing Process"
F lower, Linda S., and John R. Hayes. "Problem-Solving Strategies and the Writing Process." College English 39.4 (Dec. 1977): 449-461.
One of the earliest collaborative precursors to Flower and Hayes' cognitive process model, this essay presents heuristics and problem-solving as ways to address the morass of the same old same old: "our basic methods of teaching writing are the same ones English academic were using in the seventeenth century" (449). Flower and Hayes argue that while practitioners do well to get students analyzing writing-as-product, there has not been especially much to illuminate aspects of process. Process, they write, has been left "up to inspiration" (449). They invoke heuristics much in the same light Janice Lauer does in 1970 with her brief essay and related bibliography, arguing that heuristics are "a kind of shorthand for cognitive operations...they give the writer self-conscious access to some of the thinking techniques that normally constitute 'inspiration.'" (452). If we were use a line chart to illustrate a continuum between algorithmic approaches to invention and aleatory approaches to invention, with heuristics as an intervening term, Flower and Hayes' variety of heuristics would probably be located herex, nestled among the algorithm-heuristic hybrids. Their cognitive process model--circa 1980--resides in this same area, an area regarded with suspicion by many of the critics of the cognitive process, so suspiciously, in fact, that doubters tend to glide across it like they would Lock Ness:
<--algorithm-------x----heuristic------------aleatory-->
Still, Flower and Hayes make a positive argument, sharing their goal-directed heuristics as "an alternative to trial and error" (450) and as an approach that views writing as a "thinking problem, rather than an arrangement problem" (450). The thick teleology governing this approach is something of a concern to Berthoff (for all of the reasons she points out in "The Problem with Problem-Solving." Flower and Hayes offer heuristics as a fourth alternative to the three pervasive strategies for writing: 1.) formulism and prescription (as comes, oftentimes, from text books), 2.) inspiration (kept mysterious and often following Romantic misconceptions), and 3.) writer's block (nothing works). Heuristics offer "problem-solving techniques." Like the journalist's "Who? What? When? Where? Why?" (451), heuristics "give the writer a repertory of alternatives and the power of choice" (452). Their problem-solving strategy is derived from protocol analysis; researchers focused on two key tasks: "(1) to generate ideas in language and then (2) to construct those ideas into a written structure adapted to the needs of a reader and the goals of the writer" (452). Consider this a moment where the proposition takes on a mechanistic character. The heuristics are broken down as follows. Each item includes an explanation:
1. Plan
1.1 Set Up a Goal
1.2 Find Operators
2. Generating Ideas in Words
2.1 Play Your Thoughts
2.1.1 Stage a Scenario
2.1.2 Play Out an Analogy
2.1.3 Rest and Incubate
2.2 Push Your Ideas
2.2.1 Find a Cue Word or Rich Bit
2.2.2 Nutshell Your Ideas and Teach Them
2.2.3 Tree Your Ideas
2.2.4 Test Your Writing Against Your Own Editor
3. Constructing For An Audience
3.1 Ends
3.1.1 Identify a Mutual End You and the Reader Share
3.1.2 Decide on Your Own Specific Ends
3.2 Roadblocks
3.3 Means
3.3.1 Develop a Rhetorical Strategy
3.3.2 Test Your Rhetorical Strategy
Flower and Hayes anticipate and answer concerns about the ordering of the heuristics: "Do writers dutifully Plan, Generate, Construct, then turn out the light with the paper done? The answer is an emphatic no. Although we have grouped these heuristics together by their function, the process of writing rarely if ever exhibits those autonomous stages textbooks describe as Gather Information, Outline, and Write. Instead, thought in writing moves in a series of non-linear jumps from one problem and procedure to another" (460). They go on to call the process "iterative," but perhaps there isn't enough here too address the ways a goal changes or the sort of writing that sets out toward a moving or undetermined end. Continuing in the spirit of positive assertions, Flower and Hayes describe writing, "like problem-solving thinking in general, [as] a performance art" (461), and they are explicitly interested in "replacing the mystique of talent and the fear of failing with the possibility of an attainable goal" (461).
"Because inspiration is always dependent on the mental preparation that went before, it often does fail for the passively expectant writer waiting for the flow of magic ideas" (451).
"In formulating our strategy in this two-part way, we have made a fundamental assumption about the composing process: namely, that it can often be divided into two complimentary but semi-autonomous processes, which we designate as generating versus constructing on one level and playing versus pushing on another" (452).
Terms: product (449), process (449), heuristics (450), problem solving as hot area in cognitive science (450), protocol analysis (451), inspiration (451), prescription (451), writer's block (451), repertory of alternatives (452), power of choice (452), operators (453), goal-directed play (454), synectics (455), pockets of knowledge (455), flow(456), rich bits (456), code words (456), nutshelling (456), reverse outlining (456), pattern and discovery process (459), writing as performance art (461).
