A few posts back I mentioned Patrick Wilson’s 1983 book Second-Hand Knowledge [link], in which he argues that librarians ought to pay more attention to cognitive authority. I started writing a book review last week but I quickly realized that Wilson’s discussion is actually pretty weak. I mean, for a guy with a PhD in philosophy from Berkeley, it’s surprising how philosophically sloppy and under-researched his arguments are. But, there are a couple of interesting positions he takes and I’d like to quickly outline one that I think might be a bit polarizing.
The general argument of the book goes something like this:
- Most of what we believe comes from the testimony of other people (which includes texts, video, etc.)
- But, we don’t count all information sources as equally reliable: “some people know what they are talking about, others do not. Those who do are my cognitive authorities.” (p. 13).
- Cognitive authorities can be defined in terms of a social relationship in which one person has epistemic influence over another person with respect to some sphere of interest.
- There is a “knowledge industry” created in part to regulate cognitive authority. This includes formal institutions like publishers, universities, academic societies, and libraries that help regulate the social relationship of cognitive authority. It also includes informal theoretical systems that determine spheres of interest. These informal systems can be seen in the way intellectual fashions change over time (e.g., New Criticism vs. structuralism vs. post-structuralism vs. deconstructionism…each has its own criteria for authority).
- Libraries are a part of the knowledge industry that regulates cognitive authority.
- So, librarians should understand cognitive authority and their relationship to it.
It takes a while for Wilson to address libraries and librarians, but in Chapter 6 he turns his attention to the role of the library in the knowledge industry and he reflects on why people use libraries in the first place: they want information. But not just any information. They don’t want misinformation. They want quality information from cognitive authorities. But, given that libraries are literally filled with misinformation, there seems to be a need for some sort of quality control either at the point of collection or the point of access. Ideally, there should be someone to help information seekers determine if they’ve got the best available information. Wilson asks, “can those professionally responsible for information storage and retrieval act as quality controllers?” (p. 171).* In other words, what makes librarians trustworthy sources of information? Well, there are a few options.
First, it would seem to be the case that in order to effectively evaluate information, we ought to be experts on the relevant subject area. So, if a student comes to the reference desk looking for articles on Aztec funerary practices, I need to be an expert on Aztec funerary practices in order to identify which articles are the best. And so it goes for any subject area: a science librarian must be at least as much an authority on scientific matters as a practicing scientist, a medical librarian must be equal in expertise to a medical doctor. Occasionally you’ll even hear librarians (or, more typically their administrators) talk about hiring more PhDs to fill subject librarian lines: “we need experts.”
The only problem is that outside of the field of library science itself it’s impossible for a librarian to have authoritative expertise on anything but a very small aspect of a library collection. We hire ‘science’ librarians and ‘medical’ librarians, not ‘organometallic chemistry’ librarians and ‘cardiology’ librarians. Even a librarian with a PhD in a given field is only going to have expertise in certain areas of that field; the PhD is a mark of specialization, not omniscience. Put simply, librarians can’t be expected to be polymaths.**
However, even if we lack subject-expertise, we may have some other expertise. Maybe, Wilson suggests, librarians are “authorities on authority.” Maybe the librarian is the person who “can be trusted to tell us who else can be trusted” (p.179). We don’t have to be experts in the fields in which we can identify authorities; we just need some way determining who deserves to be taken as having cognitive authority. Sort of a meta-level evaluation of information. This certainly seems a compelling possibility, and it does lend credence to our insistence on spreading the gospel of information literacy. But, Wilson makes an interesting argument on this point. If a librarian isn’t a subject expert, all she can use are “indirect tests” of authority. These include asking
- What is the present reputation of the author of this information? (p. 166)
- Who is the publisher? (p. 168)
- Is the information intrinsically plausible? (p. 169)
Here, Wilson has crafted the beginnings of what would later develop into information literacy (even looks a little like the CRAP test doesn’t it?). But, Wilson is quick to point out that these indirect tests are something that almost any person can master. If librarians’ judgments about information quality “are based not on expertise in the subject matter concerned but only on external signs and clues, then they are based on the same sorts of things that any other person ignorant of the subject matter would have to use” (p. 181). So, librarians can’t claim some special expertise or credibility when it comes to evaluating information. There are no trade secrets. So, even if we try to elevate information literacy as the locus of our expertise, we fail.
And here we get to the reason I wrote this post: the possibly polarizing position.
If Wilson is right that librarians are not cognitive authorities on anything other than library science itself, then why do information-seekers trust librarians? The answer is not that librarians are specialists. Quite the contrary. Librarians are delegates. It isn’t that librarians are better than average at making decisions about cognitive authority, it’s that they are no worse and so people trust librarians to work on their behalf (p. 186).
Let that sink in for a moment.
Librarians love arguing their roles in their communities. Are we activists? Educators? Gatekeepers? And we love arguing about the lack of rigor in library school programs.*** Maybe we ought to stop beating ourselves up over what intellectual, political, or moral mission makes us different from the communities we serve. Maybe we just are our communities? In a certain sense, this is liberating; we can learn to evade the detachment that characterizes our profession. We can meet our communities as equals, not experts. We can understand the reasons that motivate movements like New Librarianship or critical librarianship. Wilson was on to something.
Then again, what do we lose as delegates? Probably not our professional stature: we’d still be authorities/experts on library science. But, perhaps our gravity outside of library science? The librarian is a cultural archetype and we are often called-on to weigh-in on non-library issues. Perhaps some of our advocacy? The delegate view would completely invalidate many ALA resolutions as being outside a far narrower conception of our expertise; as Wittgenstein said, “whereof one cannot speak, thereof one must be silent.” Perhaps our commitment to intellectual freedom? After all, we’d be responsible for following community opinion, even if that opinion lends itself towards intellectual conservatism. Perhaps our value as an information resource? Wilson certainly didn’t anticipate the Google age. Perhaps whatever professional pride we have left? It’s hard to say. But it’s worth thinking about. I’ll concede that this post barely scratches the surface and I hope someone else is inspired to investigate.
Summing up: Are librarians authorities on information? Are we experts on information literacy? Wilson’s argument suggests that no, we aren’t. We’re delegates appointed by our communities. I highly recommend reading Wilson’s Second-Hand Knowledge. Like I said, most of it is shoddy philosophy. But there are a few important insights. Personally, I’m not convinced by Wilson’s librarian-as-delegate argument. I’ve covered the paradox of authority and expertise in the past [one, two, three] and I reached a very different conclusion from Wilson, one in support of librarians as cognitive authorities. But, Wilson’s argument shouldn’t be discounted. Take it on my authority.
* Of course, librarians have a standard response when asked to provide quality control: evaluation requires subject expertise and librarians only have expertise in information handling and librarianship (p. 173). So, librarians have to be neutral, which is a deeply problematic position to take. And impossible to boot.
** Not to say that there aren’t librarians who are expert authorities on certain topics. There certainly are. But, professionally speaking, requiring librarians to be authorities on entire fields or entire collections is like asking for unicorns.
*** Personally, I think that programs that focus more on information science can and often do have intellectually challenging and engaging classes.
Lane, I also disagree very much with Wilson. But his argument appeals to librarians in some ways.
Wilson concludes that librarians are pure intermediaries or delegates between library patrons and information. This sort of purity is appealing to librarians. It marks us as special.
Wilson also argues, remarkably, that having a graduate degree in a subject does not suffice to give librarians any cognitive authority in that subject (p. 187). In the librarian role, according to Wilson, a librarian is a sort of cognitive ascetic, who takes a kind of vow of intellectual poverty. So Wilson describes librarians as professional skeptics.
This also may be appealing to librarians. We’re off the hook about having to know much about any subject, and yet we have some status.
There’s something attractive in Wilson’s stance. I think he’s serious about it. But this skepticism strikes me as a dead end.
In particular, Wilson says:
“Thus evaluation falls out as a problem of no interest – because evaluation is not the aim, because the standards are supplied by others, or because there is no hope of doing anything to change standards of evaluation over the opposition of others, and no need to change them if they are already approved by others.” (p. 193)
So Wilson’s skepticism means denying that librarians (in their librarian roles) are capable of evaluating information. This would seem to imply that librarians cannot teach others how to evaluate information.
Since Wilson’s view implies this, it is important to see how Wilson’s view is attractive to librarians.
I don’t think he actually denies that librarians can teach others how to evaluate information; to me it seems more like he’s arguing that anyone can teach information evaluation. Librarians don’t have the monopoly. Overall though, I agree that his skepticism is a dead end.
Maybe here’s a place also to say I very much like the idea of librarianship as a general expertise about the chain of testimony. I would have said that librarians are experts at research, but your formula is more informative. So thank you. I’m going to quote you, Lane.
Still, it seems to me that evaluation (of information) in the sense we want also requires learning at least something (just a little) from the really good testimony librarians discover using expertise over the chains of testimony.
The way I’d put is that a librarian might not have expertise over a domain of knowledge, but she is entitled to “borrow” expertise from a really reliable source by applying the standards of evaluation she learns in such sources.
For example, if all textbooks in molecular biology (and other authoritative sources) say that biologists know that archaea are phylogenically distinct from bacteria on the basis of ribosomal RNA, even a non-expert librarian is capable of knowing that this is the evidence on which the conclusion is based, and not just to conclude it is true on the basis of expert testimony. In this particular case, the evidence is simple enough for an educated non-expert librarian to understand, when it is taken on its own. So it is different from some complex reasoning in math, physics, economics, or climate modeling in this respect.
Thanks! And with your ribosomal RNA example, I’m tempted to say that the knowledge in question is still testimonial knowledge because the non-expert librarian didn’t run the actual tests in the lab. I think I lean towards the position that testimony is the most common and most important grounds for knowledge.
I have to suppose that your view about librarianship as knowledge of testimonial chains means that librarians investigate the extent to which testimony is reliable or not. So I suppose you are disinclined to give testimony the status of a priori justification.
If so, I agree with you (and Hume).
In connection with the RNA example, a large portion of the process of inference from evidence can be done by any layperson using the open access nucleotide databases at NCBI and the BLAST program. That’s not testimony. The nucleotide sequences accessible in that way come from laboratories and this might be construed as testimony from the scientists who run the labs, as you say. Still, the processes for gathering this information are highly automated and institutionalized, which greatly stretches the notion of “testimony,” in my view, taking it pretty far from how testimony is discussed in epistemology.
This is an example of why I’m uneasy (understatement) about the way “testimony” is used in epistemology as covering what may be very disparate processes. However, I suppose a defense of the expansive notion of “testimony” in epistemology might start from the way epistemology is a very general inquiry.
It’s the need for some broad view of the aims of information literacy instruction and librarianship that got us from librarianship into epistemology in the first place.
[…] A few posts back I mentioned Patrick Wilson's 1983 book Second-Hand Knowledge [link], in which he argues that librarians ought to pay more attention to cognitive authority. I started writing a book… […]
Thanks for this article, this brings ups some interesting points that I hope to explore in more detail in the short future for my own studies. I think Wilson’s statement has some validity to a certain extent, “But not just any information. They don’t want misinformation. They want quality information from cognitive authorities. But, given that libraries are literally filled with misinformation, there seems to be a need for some sort of quality control either at the point of collection or the point of access.” My response would not be so much focused on librarians as subject authorities serving the reference desk per se, but possibly a critical rethinking of the role of cataloging and classification. Once an “informational” work is classified, seldom do we retrospectively re-consider its bibliographic record unless we are ready to weed the item. For example, when the author of Three cups of tea was exposed as fraudulent and his foundation money maker was shut down, did any of our librarians stop to consider it’s bibliographic record? No one in our system seem to understand the quandry, let alone question whether our profession might need to address potential ethical issues regarding its information record. Ten years from now, people will forget about the fraudelent news; but they will still find his “nonfiction” work in the social studies classification for future readers. I believe these kinds of questions ought to be asked by the profession.
Laura
In Wilson’s view, librarians are professionally incapable of performing the evaluative tasks needed to distinguish information from misinformation. That’s where Wilson runs into a skeptical dead end.
Wilkinson and I disagree with Wilson, but for different reasons. Wilkinson, I believe, considers librarianship methods in themselves as sufficient for many or most evaluative tasks. I think that even if that’s true and I’m not so sure, librarians are capable of going further. Librarians can, at least sometimes, look at the same evidence experts look at, and that’s different from reliance on expert testimony.
Wilkinson’s idea of librarianship is that it is knowledge of testimonial chains. Wilson (in his 1968 book, Two Kinds of Power) distinguishes two kinds of bibliographic control: exploitative control (which is the ability to find the best text for one’s purposes and thus involves evaluative knowledge) and descriptive control, or the ability to arrange a corpus of texts according to some description. Descriptive control is roughly what today we’d say can be accomplished by databases, catalogs, bibliometrics, and other librarianship methods. Wilson (Two Kinds of Power, p. 35) says that reporting on the opinions of experts on a piece of writing is not to evaluate it in the sense of exercising exploitative control, but is just an exercise of descriptive control.
Wilkinson’s idea of librarianship, again, is that it is knowledge of testimonial chains. So librarianship methods, as Wilkinson understands them, are the same as the methods for what Wilson calls descriptive control.
Wilson argues (pp. 29-34) that we cannot accomplish the tasks of exploitative control (including evaluation) using descriptive control. The argument is weak, although Wilson fills in more detail in the rest of the book. A short version of the argument (p. 33) is that even an ideally perfect description of a text cannot anticipate every aim that it might serve and how well it would serve it.
So Wilkinson and Wilson seem to agree on what librarianship is about. They disagree on how much can be accomplished solely by using librarianship’s descriptive methods. They disagree on whether these methods suffice for evaluative purposes.
Slightly less abstractly, in bibliometrics, we exercise descriptive control by counting citations, and get knowledge of a kind about information: Article X has Y citations (in database Z). This may be understood as testimony from the citing authors that article X has a certain level of quality. (In my view, the concept of “testimony” is being stretched here, but that’s okay.) We might infer from this testimony that article X probably has a certain level of quality. That’s an indirect approach to evaluation. In Wilkinson’s view, most of the time this is perfectly adequate for evaluation.
This is right to some extent, but I’m not satisfied. Citations have different value in different fields and normalization of citation counts across disciplines is a puzzle (despite repeated claims to have solved it). The notion of an adequate sample of expert opinion is also a puzzle (except where there is consensus). The clichés, “Scientists say” and “A study finds” are often mocked as unhelpful or misleading. Reliance on authority is necessary at times, but it is often irrational. Indirect evaluation is weak, and just goes through the motions of evaluation, like “kicking the tires” on a car.
Like me, Wilson is dissatisfied with descriptive methods as a replacement for more substantive evaluation. But Wilson’s conclusion (in his later book, Second Hand Knowledge) is a skeptical dead end: Librarians are restricted to descriptive methods and thus cannot really know whether one text is the best for a purpose or not.
If instead we look at the evidence or reasons for ourselves, we have a more direct evaluation of the conclusions drawn in an article. If we read Euclid’s proof of the infinitude of primes and understand it, then we can see for ourselves that it is sound, regardless of how often it is cited, which probably is not very often. Descriptive methods might get us an approved translation of Euclid, not really a good idea of whether the proof is sound (that is, adequate for finding that the conclusion is true). But when we work on understanding the proof, maybe we aren’t engaged in librarianship, and instead we’ve shifted into mathematics. That does not bother me. The purpose is to find the truth and knowledge of it. This particular proof is one that any reasonably intelligent person can understand. Other examples can be drawn from other sciences.
When we rely on testimony, we can check the reliability of testimony.
So I argue that librarians are no different from anyone else in being able to learn. So information literacy instruction has no reason to be restricted to the descriptive methods of librarianship. More substantive evaluation is possible in information literacy.
Aaron: I think that’s a fair description of my position, at least insofar as I have trouble figuring it out myself 🙂
There’s a lot to unpack in your comment, and I suppose I need to read the earlier work by Wilson to respond appropriately. But, I would like to say something about your last point. I agree: information literacy has no reason to be restricted to descriptive librarianship. I’ll go as far as to say that information literacy has no reason to be restricted to librarians in the first place.
Laura, thanks for the thoughtful comment. Your Three Cups of Tea example reminds me of the Million Little Pieces fiasco a few years back. Interestingly, there is no a note in WorldCat for Million Little Pieces that acknowledges it as a fraudulent memoir.
Ultimately, though, I don’t think librarians can be held accountable for identifying every fraud or fabrication in the scholarly/literary record. Nor do I think we should be held so accountable. Though I haven’t really thought it through all the way, I’m thinking that the scholarly/literary record itself provides the necessary checks and balances. Even if a library record does not note that Three Cups of Tea is a fraud, and even if it’s place in the catalog is misleading, there are other information sources in the library that show the book to be a fraud. I guess what I’m getting at is that the place and description of any particular book or article in a library collection should always be understood as provisional and not authoritative. Other than a few special collections (such as local history perhaps) the library does not bestow authoritativeness or truthfulness on a book/article…the scholarly and literary record does (what I call the social transcript).
Anyway, it’s a really interesting point and I’ll have to think on it some more. Perhaps I’ll blog it soon?
Yes I agree, we really can’t take on such a task for every situation. That’s a daunting concept. As librarians struggle with their role as trusted delegates or community point persons, these issues with authenticity and veracity of sources will continue. As a profession, it is good to continue tackling these ethical issues as fact-checking declines and self publishing increases.
[…] hacker Topher Lawton shared this article recently. It has some interesting perspectives about our role in communities and as […]