Archive for June 19th, 2014

Threshold of the house (1)

Over the past four months, an ACRL task force dedicated to rewriting the Information Literacy standards has been releasing successive draft versions of a new Framework  for Information Literacy in Higher Education. In a nutshell, the task force is proposing a new way of looking at information literacy that moves away from the preexisting standards and towards a series of “threshold concepts.” According to the task force, these threshold concepts are “ideas in any discipline that are passageways or portals to enlarged understanding or ways of thinking and practicing within that discipline” (Draft 2, June 2014, lines 25-27). Erik Meyer and Ray Land, who developed the theory of threshold concepts, explain that “a threshold concept is likely to be

  1. “Transformative, in that once understood, its potential effect on student learning and behaviour, is to occasion a significant shift in the perception of a subject…
  2. “Probably irreversible in that the change of perspective occasioned by acquisition of a threshold concept is unlikely to be forgotten, or will be unlearned only by considerable effort…
  3. Integrative; that is, it exposes the previously hidden interrelatedness of something…
  4. “Possibly often (though not necessarily) bounded in that any conceptual space will have terminal frontiers, bordering with thresholds into new conceptual areas…
  5. “Potentially (and possibly inherently) troublesome [defined as counterintuitive, alien, or incoherent]” (Meyer and Land, 2003, pp. 5-7).

The six transformative, irreversible, integrative, bounded, troublesome concepts identified by the ACRL task force are as follows:

  1. Scholarship is a Conversation
  2. Research as Inquiry
  3. Authority is Contextual and Constructed
  4. Format as a Process
  5. Searching as Exploration
  6. Information has Value

These six “threshold” concepts are meant to serve as the anchors for the new framework and each is accompanied by a series of knowledge practices and dispositions to guide librarians as they develop information literacy curricula, reach out to campus partners, and integrate information literacy into student curricula at a variety of levels. These aren’t prescriptive standards like we’ve had, they are “interconnected core concepts, with flexible options for implementation” (Draft 2, lines 23-24). I suppose the simplest way to understand the change is to think of the previous standards as the authoritarian gym coach yelling “here are the five things you need to be information literate–learn them” the new standards are more like the hippie English teacher saying, “hey guys, here’s some stuff to think about, but interpret it whatever way works best for you.”


You really ought to read the draft, then come back. I’ll wait.

Anyway, the ACRL task force has been putting the new framework in front of us for about four months now and the response from the academic library community has been a resounding…well, take a look at some of the blog posts covering the new Framework (lightly annotated):

[These are just the ones I’ve come across and bookmarked; I’m sure there are others.]

Most librarians are enthusiastic that threshold concepts are a step in the right direction, though they would like the individual concepts tidied up a little. Jargon is a frequent concern, though, to the task force’s credit, the emphasis on metaliteracy has been toned down in subsequent drafts. Overall, there seems to be a sort of guarded optimism that threshold concepts are a welcome step forward for information literacy and a worthy successor to the previous standards.

And that worries me.

You see, I was actually on the task force until this past February; I resigned the day before the first draft came out. Part of the reason I left the task force was because I didn’t have any faith in threshold concepts as the foundations for information literacy (fun fact: I even wrote one of the threshold concepts in the current draft). The six concepts identified by the ACRL task force as threshold concepts are all interesting and I tend to agree that they are important concepts, so I’m not saying that I disagree with things like the iterative nature of research, that authority is contextual, or that information has value. Rather, it’s the…um…thresholdiness that I have trouble with. I think today I’ll offer the rare critical take on threshold concepts in general and, in coming weeks, offer further thoughts on each of the six concepts of the framework. So, leaving aside any criticism of particular concepts singled out by the task force, let’s look at threshold concepts in general.

North-South Korean border (6647230281)

The problem with thresholds

I suggested just a few sentences ago that critical analysis of threshold concepts (TCs) is “rare.” That’s not hyperbole. Since TCs hit the scene 11 years ago, the number of articles and books speaking out in favor of them has multiplied to the point that I can’t keep up. Most of Meyer and Land’s papers have several hundred citations; Google Scholar returns thousands of articles. But, searching for criticism yields only a handful of sources. In fact, I’ve only found four critical analyses: Sarah Barradell (2013), Rod O’Donnell (2009, 2010), and Darrell Rowbottom (2007). Each of these authors admits that the threshold concepts hypothesis has some kernel of truth, but that there are serious difficulties plaguing how TCs are formulated. We can break the criticisms down the following way:

1. How can probable characteristics be defining characteristics?

Both Rowbottom (2007) and O’Donnell (2010) have noted that threshold concepts are defined in terms of the weakest possible modalities. Meyer and Land tell us that threshold concepts are “likely to be…probably irreversible…possibly often (though not necessarily) bounded…potentially (and possibly inherently) troublesome” and so on. These hedges are concerning because they force the question of whether a putative threshold concept is actually a threshold concept. So, while there is some truth to the concept that “scholarship is a conversation” it’s trivially true that it’s a threshold concept on Meyer and Land’s schema because of course it’s possible bounded, potentially troubling, and so on. Really, the hedging language just opens the door to any concept as a candidate threshold concept. Here, I’ll make one up right now: Keywords are Perspectives. This is the idea that different perspectives on a topic may understand the topic in different terms. Sort of like how searching for “global warming” will return different results from “climate change.” Is this a threshold concept? Well, it’s possibly transformative, potentially troublesome, might be bounded, could be integrative, and I bet someone might find it irreversible, so, by definition, it is! Yay me! I just added to the framework!

But wait, it gets better. Because the five characteristics of threshold concepts are only possible characteristics, there could be threshold concepts that lack all five characteristics. Or that only have one, two, three, or four characteristics. Really, any combination works, which, if you recall your modal logic, means that literally every concept is a threshold concept, no matter how trivial, how incoherent, how false. Access Articles through the Databases? Threshold concept. Books are our Friends? Threshold concept. Libraries Jump Spanish Sandwiches? Threshold concept.

2. Concepts do not imply abilities

Rowbottom points out that the definition of threshold concepts equivocates over the term ‘concept’. Basically, there are two primary ways of understanding what a concept is. First, a concept is sometimes defined as a mental representation of something, i.e., a mental model in our language of thought. Like, I have the concept of ‘book’ because I have a word-like symbol for ‘book’ in my language. Second, some define a concept as an ability to think of, classify, or recognize something. Like, I have the concept of ‘book’ because I have the ability to recognize a book when I see it. (Rowbottom mentions a third, Fregean view of concepts, but we don’t need to be concerned with it). Rowbottom’s insight is that which view of concepts we hold will affect our method of determining whether a student has mastered a given concept. He gives the example of knowing how to play tennis versus being able to play tennis. I mean, I know how to play tennis, so I have that concept in the “mental representation” sense. But, I’ve never actually played tennis, so I would seem to lack evidence of the concept in the second sense of “concept”.

This observation raises an interesting problem for the Framework. When teaching the threshold concept “Scholarship is a Conversation” are we supposed to teach each knowledge practice in order for the threshold concept to be met? Could a student be able to perform each of the six knowledge practices without understanding the overarching concept? (It seems that way). Is proficiency in each knowledge practice a prerequisite for acquiring the concept? The basic point I’m trying to make is that the connection between having a particular threshold concept and having certain abilities is nebulous at best, nonexistent at worst. Moreover, if a threshold concept is reducible to knowledge practices qua abilities, then aren’t we back to the old standards? Aren’t we just listing skills that the information literate student must have?

Actually, it’s more complicated than that. If the concepts are reducible to specific abilities and dispositions, then the question is whether the individual abilities/dispositions are what’s transformative/troublesome/etc., or if it’s the overarching concept that’s transformative/troublesome/etc. Are the six abilities listed as knowledge practices under Scholarship is a Conversation what’s troublesome? Is summarizing changes in scholar perspective over time transformative and what not? It doesn’t seem that way. In fact, it seems as though the skills and dispositions are fairly straightforward. But, if the skills and dispositions aren’t the troublesome/transformative parts, then what is? What is there in the overarching concept Scholarship is Conversation that constitutes a threshold, independent of the listed skills and dispositions?

3. Being troublesome or transformative are agent-relative properties

Rowbottom and O’Donnell both point out that a core problem for threshold concepts is that they are agent-relative: what is transformative for me might not be transformative for you. What is troublesome for you might not be troublesome for me. So, whereas the concept “Format is a Process” might not be troublesome for me, because I learned how to research just before the advent of online search, you may have more difficulty because you are only familiar with online research and don’t understand the differences. Likewise, we can ask how transformative and how troublesome must something be to count as a threshold, which brings us back to the first objection about all the hedging language. The key thing here is that threshold concepts have a way of reducing all of our students to a single idealized student who learns a particular way. But, we know that isn’t the case. In a room of 30 students, each student will have a different standard for how troublesome or transformative a concept is. Tell your first-year students that Scholarship is a Conversation and some will get it instantly, some will struggle, some won’t get it at all. Does that mean that Scholarship is a Conversation is only a threshold concept for some students? Hard to say.

4. Do disciplines really have a unified body of knowledge?

O’Donnell (2010) raises what I feel is the most damning criticism: that the threshold concept hypothesis requires us to reduce disciplines down to core sets of unchanging beliefs. The push to have students “think like an x” (a doctor, an engineer, an economist, a librarian, etc.) has negative impacts on critical thinking, O’Donnell argues, because “if we want creative thinkers and innovators, we need graduates capable of moving outside the x framework and operating within multiple frameworks” (2010, p. 9). Similarly, he argues, threshold concepts reinforce siloing and adversely impact inter-disciplinarity. I’ll give an example. Consider an engineering student being taught according to the information literacy framework. Within engineering, she may meet the threshold concept that “the semantic aspects of communication are irrelevant to the engineering problem” (Shannon, 1948, p. 1). Through the library she may encounter the threshold concept Information has Value. Will she then have two sets of contradictory, irreversible threshold concepts? It would seem so and the only way to avoid the issue is to start carving up knowledge into non-overlapping domains, keeping engineering out of information science and vice versa.

Actually, it’s worse than that. Even within a single discipline, there are often radically incompatible views held among practitioners. For example, I actually disagree that scholarship is a conversation (my next post, I suppose). Or, for example, a deconstructionist librarian would disagree that Format is Process (il n’y a pas de hors-texte). Does this mean we’re not information literate? Not librarians? If you look back at the debate Dave Lankes and I had a few years ago over social constructionism, is it the case that only one of us is a librarian? If so, which one? If not, and we’re both librarians who just disagree on a foundational concept, does that mean there are two contradictory threshold concepts: Knowledge is Constructed vs. Knowledge is Discovered?

Even worse than that is the problem Barbara Fister alluded to on 27 February (link above). If we’re going to talk about disciplines having threshold concepts, we have to ask “whose threshold concepts?” As O’Donnell argues, “the view that there is a single set of threshold concepts in a discipline typically reflects the view that a discipline only has one reputable school of thought.” (2010, p. 9). And, as I pointed out in the last paragraph, this isn’t the case: most disciplines have multiple schools of thought. To take a clear example, what would the threshold concepts in psychology be? Would they come from psychoanalysis? Behaviorism? Cognitivism? Humanism? Again, whose threshold concepts define the discipline? Ultimately, O’Donnell argues, it boils down to power and control. Whoever controls the dominant narrative decides the threshold concepts. In most cases, this reduces to majority academic opinion, I would think. That’s still a bad way to do things, but outside of kowtowing to only the most mainstream academic thought, there are deeper sociological concerns. As Fister alluded, it’s possible that the threshold concepts being pushed in the Framework are only reflective of one particular view of information literacy. Probably white, probably middle-class, probably well-educated. That is, probably people in a position of privilege. But, I don’t really know. (The threshold concepts put forth by the committee were decided upon by an anonymous group of librarians in a “Delphi study.” The task force was not privy to the names or affiliations of Delphi study participants, nor were we given any justification, evidence, research, or other reasons to accept the concepts we were given. The role of the task force was to rewrite and expand upon the concepts given by the Delphi study, not to ask for justification.)

Ultimately, the threshold concept hypothesis requires us to view disciplines as monolithic, when we know that isn’t really the case. Likewise, because threshold concepts are agent-relative, there’s the very real problem (and likelihood) that any given set of threshold concepts is but a reflection of power and privilege. The entire theory of threshold concepts has a funny way of oversimplifying the very real distinctions and difficulties that are inherent in a body of knowledge.

Rethinking threshold concepts for information literacy

So, where are we with the ACRL framework? Like I said earlier, I’m not necessarily opposed to the actual concepts listed by the task force. They seem reasonable and what problems I have with them will be addressed in future posts. But, I am troubled by the connection to an intentionally vague, conceptually muddled, agent-relative, and reductionist theory. It would be one thing if the aforementioned problems with threshold concepts had simple solutions in the literature. Like, if Meyer and Land or someone else showed how the objections are unfounded. But, that hasn’t happened. Scholarship may be a conversation, but in the case of threshold concepts the dialogue is permanently one-sided. Of the 32 articles that cite Rowbottom, not a single one actually engages with the arguments. It’s always: “some have criticized threshold concepts, but we can ignore those criticisms because threshold concepts are popular.” Or, once or twice I’ve seen hand-waving to the effect that, “sure threshold concepts are vague and reductive and sure anything can be a threshold concept, but that’s only a problem for realists, not for us postmodernists.” No one is having a conversation about the legitimacy of the threshold concept hypothesis. According to the new framework, that means there’s no scholarship to show it’s a legitimate theory.

Personally, I do think that there are troublesome, transformative concepts out there. I do think there are Gestalt shifts and “Eureka!” moments in information literacy. But I don’t think that we can identify them by fiat. I don’t think that everything that’s troublesome has to be transformative. Or that everything transformative is integrative. And so on. Really, the six concepts in the Framework are a good start and they make sense. More importantly, they can stand on their own quite independently of the threshold concept hypothesis. I don’t want to justify the value of information literacy in terms of some mysterious threshold that students must face down. I want to introduce students to the complexities of scholarship and research because it’s just good to know that stuff. We don’t need pedagogical gimmicks like thresholds to see that students would benefit from certain concepts about research.

Anyway, those are just some of my thoughts. Maybe in the next few weeks I’ll say something about each threshold concept in turn (like how I don’t think scholarship is a conversation).

Doorway Empty


 Stuff I cited

Barradell, S. (2013). The identification of threshold concepts: a review of theoretical complexities and methodological challenges. Higher Education, 65(2): 265-276.

Meyer, J. H. F. & Land, R. (2003). Threshold concepts and troublesome knowledge: linkages to ways of thinking and practising within the disciplines. In C. Rust (ed.), Improving Student Learning–Ten Years On. Oxford: Oxford Centre for Staff and Learning Development (OCSLD), 412-424.

O’Donnell, R. (2009). Threshold concepts and their relevance to economics. In T. Robinson, T. Tang, & A. Fletcher (Eds.), ATEC 2009, 14th Annual Australasian Teaching Economics Conference Proceedings. Paper presented at the Australasian Teaching Economics Conference, Brisbane, 13-14 July (pp. 190-199). Available here: http://www.atec2009.bus.qut.edu.au/documents/Binder3b.pdf

O’Donnell, R. (2010). A critique of the threshold concept hypothesis and its application to opportunity cost in economics.(Working Paper No. 164). Retrieved from http://www.finance.uts.edu.au/research/wpapers/wp164.html

Rowbottom, D. P. (2007). Demystifying threshold concepts. Journal of Philosophy of Education, 41(2): 263-270.

Read Full Post »