Off the Mark

Sunday, 30 Sep

van Rijsbergen, C. J. (1986). A new theoretical framework for information retrieval. Proceedings of the Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, 194-200. Retrieved via ACM Portal.

Cited by Hjørland (2007). Semantics and knowledge organization. ARIST 41: 370.

A useful paper in that the author declares:

I have reluctantly concluded that the fundamental basis of all previous work is wrong. Almost all of the previous work in Information Retrieval (including my own) has been based on the assumption that a formal notion of meaning is not required to solve the information retrieval problem (194).

In discussing the need for a formal semantics:

That is, a document is retrieved if it logically implies the request. However, as we all know, documents rarely imply requests; there is always a measure of uncertainty associated with such an implication. And so, a notion of probable, or approximate, implication is needed …. Modelling the information retrieval process in this way goes beyond the keyword approach, and specifies, once and for all, what relationship between a document and a request is to hold to compute probable relevance (195, emphasis mine).

This is (one big) reason why computer-based IR, as good as it may become, is doomed to incompleteness. There is simply no way, no freaking way, in which anyone could ever specify, once and for all, all of the relevance relationships between documents and a request, much less specify those formally. [But, then, human-based IR faces the same problem for but for somewhat different reasons.]

He does go on to show that he does knows a bit about relevance, such as documents themselves are not, in fact, relevant to requests. And one must love the wonderfully named Logical Uncertainty Principal, which is the main product of this paper.

Peregrin, J. (2004). Pragmatism & semantics. English version of Pragmatism und Semantik. In A. Fuhrmann & E. J. Olsson (Eds.), Pragmatisch denken (pp. 89-108). Frankfurt am Main: Ontos. English version retrieved 30 Sep 2007, from http://jarda.peregrin.cz/mybibl/PDFTxt/482.pdf.

Cited by Hjørland (2007). Semantics and knowledge organization. ARIST 41: 372.

Discusses what he calls the Carnapian and Deweyan paradigms in language. The intent is to show how “the technical apparatus engendered by the Carnapian approach, with is wealth of results, can be put into the service of the Deweyan paradigm - if we liberate it from the Carnapian representationalist ideology” (3).

On Wittgenstein’s analogy to chess:

Thus the meaning of an expression can be compared to the role of a chess piece, which acquires its role of, say, a ‘knight’ by being handled according to the rules of chess (4).

But meanings and rules can be played upon; are these just alternate rules, or mis-use of the rules to another end?

Makes us of Sellars’ rules of semantics as rules of inference, which relies heavily on the primacy of sentences and on locating sentences in a logical space as propositions. But it simply is not the case that any of the bits below the sentence level have no meaning, nor that communication can not occur with sentence fragments or single words.

And the whole logical space/proposition issue is heavily positivistic! Clearly not all communication is propositional.

Such objections point out that if we start to treat formal semantics as the basis for a philosophy of language, we are likely to run into a vicious circle: we reduce philosophically problematical concepts to the seemingly perspicuous formal semantic concepts, which, however, ultimately rest on the obscure concepts to be explicated (10-11).

Amen to that!

But to place the Carnapian approach in the service of the Deweyan he falls back on possible world semantics. Gah! Can we please do away with the so-called possible worlds? Possible worlds are an supra-metalinguistic way of talking about our already common-sense, lay, metalinguistic way of discussing alternative scenarios and logical possibility and necessity. To formalize this way of talking into possible world semantics leads one easily down the path from a linguistic way of knowing (epistemology) to postulating actually existent possible worlds (metaphysics).

On the pragmatic fallback, as I am tentatively calling it (them?):

And I think that the inferentialist should realize that modeling is a very useful thing. Thus I think that although language is not literally a nomenclature or a code (as the Carnapian paradigm has it [orthodox linguistics]) it remains useful, at times, to see it as a code, just as it is often useful to see atoms as cores orbited by electrons (12).

This is a very interesting paper, but I do not think it has won me over to its way of thinking. I am concerned that we will, especially in IR, have to resort to the pragmatic fallback. But Sellars’ view is still far too positivistic and thus rules out much of what we would call communication. Perhaps this view was acceptable when libraries were the gatekeepers and we dealt only in “serious” reading material. But this is, in some respects, a new age and the past age is long past. Perhaps libraries need not worry about some of this when one considers the sorts of material that they deal with (but I doubt that!). But KO and IR is much broader than libraries. And even if KO and IR uses a sub-set of our theories of language and communication (assuming we separate them; perhaps not), we should have theories that cover all of communication and language and then explicitly pull out the bits we need. We should not be starting from a limited theory to begin with.

Harris, Roy, and George Wolf, eds. Integrational Linguistics: A First Reader. 1st ed, Kidlington, Oxford, UK: Pergamon, 1998.

Re-read:

• Introduction
• Ch. 1: Harris, R. “Language as Social Interaction: Integrationalism versus Segregationalism.”
• Ch. 2: Harris, R. “The Integrationist Critique of Orthodox Linguistics.”

Read:

• Ch. 8: Harris, R. Three Models of Signification.”

I skipped ahead to chap. 8 as I want to get a handle on the integrationist view (i.v.) of meaning.

Discussion of these is going to have to wait.

Monday, 1 Oct

American Society for Information Science and Technology. Theories of Information Behavior. Medford, N.J: Published for the American Society for Information Science and Technology by Information Today, 2005.

Preface

• Ch. 1: Bates, Marcia J. “An Introduction to Metatheories, Theories, and Models.”
• Ch. 2: Dervin, Brenda. “What Methodology Does to Theory: Sense-Making Methodology as Exemplar.”
• Ch. 3: Wilson, T. D. “Evolution in Information Behavior Modeling: Wilson’s Model.”
• Theory 60: Hjørland, Birger. “The Socio-Cognitive Theory of Users Situated in Specific Contexts and Domains.”
• Theory 2: Belkin, Nicholas J. “Anomalous State of Knowledge.”
• Theory 5: Bates, Marcia J. “Berrypicking.”

This book looks useful enough that I ordered my own copy, with my ASIST discount of course. If you have the slightest aversion with authors referring to themselves in the third-person or heavily self-citing then you may want to skip it or take it in small doses. But the self-citation in many cases makes perfect sense as many of the authors are writing about their own theories. But the third-person stuff, especially the “Article x is clearly a most influential paper in LIS having been cited 642 times” [made up example], is simply past precious.

The book as a physical item seems to be of fairly good quality, although I do have a few gripes. Page margins are far too limited, especially the outer margins. The type face is generally readable, although a tad too small for some, but it has two features I do not like. First, and only minimally pain-inducing is the hyphen, which slants upward from left to right at about a 40 degree angle. Far worse, and especially grating since it occurs extremely frequently due to citation style and time period of most citations, is that the numeral 1 is a capital I. WTF is that? I realize that some old typewriters and perhaps early computer printers used either an “l” or an “I” for a “1″. But this book was published in 2005! Why would anyone use a type face that uses a capital I for a 1 in 2005? Information Today should be ashamed. [it also has a ridiculously long "/".]

I primarily checked this book out to get a copy of Hjørland’s “The Socio-Cognitive Theory of Users Situated in Specific Contexts and Domains.” It will also be of immense value in the section of my paper where I critique various aspects of our field. By providing a brief overview of 72 theories in a lit review format, along with highlighting applicable research projects, the book will prove exceptionally useful.

I read the above theories to try and get a handle on how they might or might not fit in with Integrationism.

Hjørland’s use of the socio-cognitive view and domain analytic theory can, I believe, easily be given an integrationist reading. Within integrationism, the “three parameters relevant to the identification of signs within the temporal continuum” are biomechanical, macrosocial and circumstantial [Harris, see previous link]. The biomechanical and macrosocial parameters are clearly shown in Hjørland and, I believe, the circumstantial can be pulled out of the “socially constructed” easily enough.

Belkin’s anomalous state of knowledge (ASK) is explicitly cognitivist and, thus, may not translate as well. It most certainly will not fall under Hjørland’s views easily. What is his view of ASK? [Note to self to ask him; noted.]

Bates’ Berrypicking; hard to say from this article. Seems as if it could fit in many other views and theories. Unfortunately, the assumptions and epistemologies underlying her model are almost completely opaque in this article. Will need to check the original articles themselves.

Schneider, K. G. “Range of Desire: In the military, I learned to love women and guns.” nerve.com

Very enjoyable read. Parts of this resonated deeply with me, some parts not so much, and some seemed very different than my experience. But this is Karen’s story so that last clause in the previous sentence isn’t too relevant.

Hjørland, Birger. (2002). “Epistemology and the Socio-Cognitive Perspective in Information Science.” JASIST 53 (4): 257-270.

Through the lens of psychology literature demonstrates the differences between the cognitive and socio-cognitive views, discusses domain analysis, shows that knowledge of subject literature(s) is required for effective info retrieval, demonstrates that different paradigms and epistemologies imply different information needs and relevance criteria.

Some of these points ought to be blatantly self-evident but they generally ignored in our literature. These points can fit within an integrationist view most likely.

Hjørland, Birger. (2004). “Domain Analysis: A Socio-Cognitive Orientation for Information Science Research.” Bulletin of the ASIST, Feb/March 2004: 17-21.

This is good, but short, overview of domain analysis based on the author’s talk at the ASIS&T 2003 Annual Meeting. For anyone looking for a short intro to domain analysis and several other of the author’s views (socio-cognitive view, pragmatic realism) this is a great place to start.

For some reason the close juxtaposition of IS & IT in the 1st several paragraphs of this article made me make an odd sort of observation:

IS and/vs. IT

“is” and/vs. “it”

being and/vs. thing

So tell me about relevance again, will you? Relationships are defined by what?

Tuesday, 2 Oct

Davis, Hayley G. Words: An Integrational Approach. Richmond, Surrey: Curzon, 2001.

• Ch. 2: Methodology: The Word of the Layperson
• Ch. 3: What Do Lay Speakers Say About Words?

Wednesday, 3 oct

Hayley (above).

• Ch. 4: Words and Linguistic Meaning

Hjørland, B. “Domain analysis in information science: Eleven approaches - traditional as well as innovative.” Journal of Documentation 58 (4), 2002: 422-462. doi: 10.1108/00220410210431136

This is a long but useful article about the uses of domain analysis in information science. It pointed me to several resources of which probably ought to play a role in my critique of language theorizing and use in LIS.

I loved this quote, under the head of Indexing and retrieving specialties, as it serves to justify my extending stay at GSLIS:

Too often library and information specialists feel they lack adequate subject knowledge. In order to claim the existence of the field as a serious field of study one has, however, to develop sufficient subject knowledge in at least one field (e.g. LIS itself). The application of LIS principles to a specific task may make research in information science more relevant and realistic (429, emphasis mine).

The following is a claim made in many places by Hjørland which I am going to need more time to formulate an adequate response to, but I want to note it here:

The tendency to try to measure users’ information needs by questioning them or by studying their behavior seems to me to be mistaken. What information is needed to solve a given problem is not primarily a psychological question, but a theoretical/philosophical one (431).

While I tend to agree with this, at least in restricted domains, I do not think it is so applicable in, say, general culture. Certainly there are assumptions I am making if I want to do a Google search on Britney’s custody woes as reported in the popular press, but I do not think theory and philosophy are going to be of much use and certainly will not be dominant in my “need.”

Thus, I am led to think that this is going to be more of a continuum, and perhaps/more likely multi-dimensionally continuous. I think Hjørland’s view on this is a bit too influenced by scientific-type knowledge, “serious research” and the academic environment. But if IS and KO only focus on these limited areas of knowledge then the game is already up. We must have a wider influence or the Googles and Microsofts of the corporate world will quickly eat us up. [Noted to ask him about this.]

His spin on bibliometrics, here and elsewhere, makes it seem like they can possibly be given a integrationist spin (e.g., p. 433).

On taking the easy way out citationally (underrepresentation and overrepresentation):

In LIS there may be a corresponding tendency to overcite easy theories and methods at the expense of more difficult but also more important papers (435). [Oh, like Bush, perhaps.]

Under Document and genre studies:

These important concepts need, however, to be based on more general theories of documents, their communicative purposes and functions, their elements and composition and their potential values in information retrieval. Different disciplines or discourse communities develop special kinds of documents as adaptations to their specific needs (437).

Seems pretty integrative and reflective of the macrosocial, and perhaps of the circumstantial as well.

Terminological studies, language for special purpose (LSP), database semantics and discourse studies was the most productive citationally for me. LSPs and sublanguages will be critical to my critique of language in LIS. Can we legitimately speak of sublanguages within Integrationism, or must they be given a different spin? LSPs seem to reflect the macrosocial at first blush.

Ammon’s sociolinguistic theory of LSPs seems useful in cross- and interdisciplinary information seeking (444-445).

Spells out Hjørland’s approach (so far) to LSPs and database semantics (4 main assumptions) (445-446):

1. “Signs and their meaning are formed by social groups primarily as part of the social division of labour in society.”
2. “Different communities develop specific document types of more or less different compositions.”
3. “The above mentioned discursive or epistemic communities are always influenced by various epistemological norms and trends, which also influence the social construction of symbolic systems, media, knowledge, meaning and semantic distances.”
4. “When documents are merged in databases information about implicit meanings from the prior contexts are lost.”

Is the concept of semantic distance tenable in Integrationism?

Under Structures and institutions in scientific communication we get an explicit comment on the “narrow” view taken by Hjørland (at least in this arena) that I critiqued above:

They do not, however, cover mass media, organisational communications, and broader communications connected to the public sphere (447).

Another comment with which I basically agree but also find somewhat narrow [although he does say "a"]:

In LIS a central goal is to provide users with information which can help evaluate the validity of different knowledge claims. To help the user establish his own views on some issue based on studies of all available arguments is extremely important in LIS (450).

What can I say, except “Read it!”

Thursday, 4 Oct

Walrod, Michael E. “Language: object or event? The integration of language and life.” In Nigel Love, Ed. Language and History: Integrationist Perspectives. London: Routledge, 2006: 71-78.

Need to copy this and re-read it as it is the selection for Metadata Roundtable Wednesday. Am I the so-called discussion leader for this one?

Thursday - Friday, 4 - 5 Oct

Hayley, (above).

• Ch. 5: Parts of Speech and Grammar
• Ch. 6: Folk Characteristics of Words (split over T/F)
• Ch. 7: Reorientation: The Integration of Speech and Writing

This was actually a quite entertaining book using an “ask-the-speaker methodology, using fieldwork and interview techniques” (ix-x) to focus “on the uses to which English speakers on the one hand, and linguistic theorists on the other, out the word word” (ix). It is also a fast read.

In fact, it was downright hilarious at points. My only complaints are that: (1) it, although very relaxed, if you will, for an academic book, is still very British in style and, (2) some of the author’s conclusions did not seem to follow from the way they were phrased in summary, although they did from the evidence. Thus, I was a tad confused at points. Well worth a read if you can get it from a library. Just don’t make it an even faster read by skipping what the informants say; that will be important to coming to the correct conclusions and are, of course, the actual funny parts.

Some of things they “blame” on Americans are downright hilarious. This is not the funniest one but one I can find at the moment:

G: . . . shit as far as I understand it is being used more and more by American young girls as an expression of disgust (152)

On epistemological and ontogenetically we get:

Other guesses were that epistemological was possibly a ‘religious’ word (B) because of the word epistle, and W thought they both sounded ‘like words the Americans have made up . . . funny words” (171).

The take home message is not, of course, the humor [perhaps I ought to write humour?] but the variability of users experience with and use of metalinguistic thinking and talk contra the linguistic theorists who think we all have the same ideas innately. Well, we clearly do not nor should it take a research project and book to demonstrate that.

Friday, 5 Oct

Hjørland, B. “Arguments for Philosophical Realism in Library and Information Science.” Library Trends 52 (3), Winter 2004: 488-506. Available in IDEALS at http://hdl.handle.net/2142/1685 [pdf]

Title reflects the paper quite well.

Empiricism is a problematic philosophy, but this does not, of course, imply that empirical research is mistaken (493).

Much more interdisciplinary work needs to be done in the philosophy of science (494).

Being a subscriber to Philosophy of Science I’d say that it is beginning to be done, and I have no doubt much more is being reported in other venues. But the point is well taken and supported by me.

… the socio-cognitive and domain analytic view assumes that “in the beginning there is a community” as well as a body of more or less substantiated knowledge claims; its distinguishing charge is to locate interactional processes in their social structural context as well as in their theoretical-substantial context (496, emphasis mine).

Sounds pretty integrational to me.

Related to an above critique of relevance:

The validity—and thus the relevance—of a document claiming that a certain substance is relevant as a cure for cancer is also ultimately decided in medical research, not by asking users of information services. [en 17, 18] Thu we have a central realist claim: A given document may be relevant to a given purpose, whether or not the user believes this to be so. [en 19] (497).

Sorry but I am not reproducing the endnotes here. While I want to concur with these statements I cannot without qualification. The ultimate question whether a specific substance is a possible cure for cancer is certainly an empirical one, but assuming that our “users of information services” are cancer researchers there is a definite sense in which the relevance of that particular document to their research program is theirs to make. They may lose a Nobel over their relevance decision if it is the wrong one, but the fact that epistemologies and assumptions imply relevance also implies that the decision of relevance is somewhat in the hand of users. But the point which I fully support is that one cannot reduce relevance entirely to what the user says is relevant. In some cases there will be an objective matter of fact of some thing’s relevance to a specific question.

Further:

It is rather a claim that relevance is not a subjective phenomenon but rather an objective one. To be engaged in how to identify what is relevant is to be engaged in scientific arguments, ultimately in epistemology (for a more detailed discussion on the realist position in relevance research, see Hjørland, 2000a and Hjørland & Sejer Christensen, 2002) (497).

Yes, perhaps it is an idealist position that some part of relevance is subjective. Nonetheless, this is the case. The first sentence in the above quote is a non-starter in that it is an either/or when it needs to be an and both. The and both will differ along a continuum depending upon the domain under investigation, but it is not one or the other. What about pop culture? Again, why with such a narrow view of KO and IR?

The field of information-seeking behavior has in a similar way been dominated by antirealist tendencies. When people seek information, they have given systems of information resources with given potentialities at their disposal (497).

OK. This is objectively the case on one description. But these given potentialities are rapidly changing, and many are not so “given” anymore. There is also the matter of knowing, and even being able to know, the given of some of these systems today. This ties directly into my stated intention to hire several librarians to help me manage all of my “systems of information resources” when I win the lottery.

Anyway, I do agree with much of what Hjørland says in this article and elsewhere. I just see some things that to me seem to be based on a narrower view than I feel we can afford to take or which need a bit of nuance as I see it.

Perhaps my views are different and perhaps seem muddled to some because I am a realist about much of the external world, but I am not a realist about much of modern science. Atoms and beyond? Not so much. Useful theoretical entities they be, but just as “wrong” as Newton’s mechanics. Who’s to say our current sub-atomic particles are truly existing entities? See, there’s the rub. I am an ontological realist (generally), but I am most certainly not an epistemological realist. In fact, my dislike of epistemological realism runs much deeper than disavowing “the view that science provides a true or realistic picture of the world” (490), especially since some would say the only true or realistic picture of the world. Nope, call me an epistemological agnostic, if you like. I think epistemology is an important subject and I fully agree with Hjørland in his claim that it is central to LIS. I just don’t think we really have much that amounts to Truth or Knowledge or, more accurately, that we can ever know if we do.

It seems my views are pretty much in accord with Hjørland’s based on endnote 24 to this article (no idea what his views on particle physics is, though). And while I do agree that our subjective knowledge can be objective, in the sense that it is “in accordance with its object” (504), I do not believe that we can ever know that is is. All we have to go on is the use that that knowledge makes for pragmatically.

I have a definite post in me about science as a belief system right now but I doubt I’ll have time to get to it. I promised a friend of mine the other day who shocked me by claiming that it was not (and says she did before) that I would write it. But, alas, probably not. Trying to claim otherwise via dictionary definitions, statements by scientists, lay views of “systems of belief,” etc. simply cannot get you out of your dilemma of belief. I read a good article somewhere in the last day or so that I wanted to ask her to read. Damn it! What was it? Was it this article or something online?

Theories of Information Behavior [see above].

• Theory 10: Rieh, Soo Young. “Cognitive Authority.”

Cognitive authority theory was developed by Patrick Wilson in his book, Second-hand Knowledge: An Inquiry into Cognitive Authority. It appears that people of many epistemological persuasions have made use of Wilson’s theory. I think cognitive authority can easily be given an integrationist reading as I can see it being definitely influenced by biomechanical, macrosocial and circumstantial parameters.

Browne, Glenda. “The Definite Article: Acknowledging ‘The’ in Index Entries,” The Indexer, vol. 22, no. 3 April 2001, pp. 119-22.

This article won the 2007 Ig Nobel Award for Literature. I saw this 1st a few days ago at 3 Quarks Daily and then a few other places. When I saw the Thingology post on it this morning I finally read it.

The Ig Nobel is given “For achievements that first make people LAUGH, then make them THINK.”

Ig Nobels at 3 Quarks Daily and at Thingology. 2007 Award Winners at the Annals of Improbable Research site.

As Tim says, “Hey, it’s a problem” and the author makes some good points.

Initial articles are the focus of my Python programming so far in LIS452. My 1st program took an internal list of mixed case titles and put them in lower case, stripped leading articles (English only) and then alphabetized them. My 2nd program which is currently beta and due Thursday does pretty much the same thing except it is written using functional vs. procedural style and it reads the titles in from a file and writes them out to a 2nd file. I hope to “fix” it to capitalize the 1st letter of each title, and if I have time to use regular expressions to do the stripping. Regex will be overkill for this program but I see them as probably the most important thing I can learn from this class (at the moment anyway).

Not sure how far I’ll get with this, though, as. must. prepare. for Dr. Hjørland’s visit this coming week!

Not going to claim that I won’t be reading or re-reading anything else today but I am going to cut this off and get back to my commentary o Hjørland’s “Semantics and Knowledge Organization” which is a much bigger job than I was thinking. It is about to become a multi-post job.

Gulp. I have 3 Downey chapters and 2 Zelle chapters to read for 452, which is LEEP on-campus this week. Luckily I have an extra day to get to those since class is Friday this week. Thank the LEEP gods for that one!