Levi Bryant has a great post on the problem with the term "construction." I think his point echoes those that Latour has been making for some time. We start with the presumed nature/culture divide of the modern world, but one that is continually unmade in issues ranging from gender, race and sexuality to environment and economics. We have the critical turn of investigating conventionally natural terms as historically constructed, but only be insisting that they are not natural at all but only discursive. I.e., there may be some "natural" characteristics somewhere behind gender (you think?) but we can only access gender as a discursive term. Constructivism, as a term, is tied up in this, with the idea that something that is constructed cannot be natural (i.e. "real").
What constructing ought to denote, but perhaps never will (hence Levi and Latour's calls for a new term), is that the knowledge we produce is another object in the world, made from other objects in the world (including us). As one object among many, the knowledge we produce does not capture/represent in some pure way other objects in the world. It isn't "true" in that sense. As academics we already accept this across the campus. However it also isn't "untrue" or operating in a separate, noncommunicating realm from other objects. It isn't purely discursive or purely social.
I was thinking about this issue in what is a very mundane context for academics: assessment. I'm on one of a dozen committees helping to prepare for our Middle States evaluation (Middle States is the accreditation agency for universities in our region). That also trickles down to my department where we are having conversations about program assessment and where we also talk about assessment within the composition program. As I see it, the underlying problem that humanists have with assessment is related to this Latourian issue. It is absolutely bizarre if you think about it. I don't believe one could function and survive as a human for very long without implicitly acknowledging that the knowledge we produce about the world is more than purely subjective and discursive. Why charge your mobile phone? Or step on the brakes in your car? Or tap away at a keyboard? Why would one eat food or not try flying from rooftops if one believed that knowledge was purely subjective? Or does one insist that ideology creates the force of gravity? Of course not. No one thinks these things. We simply think that what we do know about the world isn't the world. The difference is subtle and we sometimes fail to see the subtlety. No one believes that we understand what gravity "really" is, but we do believe that our understanding of gravity is real enough for us to take care in high places.
And in my estimation, that's what human knowledge is generally about: not creating a mirror of a real world but creating real enough knowledge objects. Enough for what? I suppose those purposes can change.
But let me return to assessment. Most humanists I have encountered object to assessment because they don't believe the knowledge assessment produces is "real." There is no doubt that assessment knowledge is constructed. In fact, there are elaborate tools and procedures that are involved in producing assessment. However, this does not make assessment less real. Instead, like Latour's story of the scientist who takes the skeptic through his lab one procedure at a time, with assessment one can follow the path and investigate the actor-objects through which knowledge is mediated. As such, the knowledge that is produced connects in a different way with other knowledge. That is, if we follow the same procedures at different sites and/or at different times, the knowledge objects we produce at those different times and places has a stronger relation with one another. It doesn't mean that knowledge is perfect.
The alternative is a kind of anecdotal sharing. And I am fully in support of talking about teaching! As one of my grad school mentors, Steve North, discussed long ago, the "lore" of the hallway and office is one of the central sites of teaching knowledge. Lore has its own kind of networks, its own constructedness. However, lore has a different relationship than assessment to other (knowledge) objects. Conventionally we might say that when we shift from lore to scholarship about teaching that this is a purely discursive shift or that it is about social-power relations. These things are partly true I think, but they are only part of the story. The other part is that research is constructed differently and thus has different strengths in its mediation of network relations, and this construction is not "purely" discursive. It has to do with the world of objects as well.
So, for example, we might anecdotally say that our undergrads are good at the close reading of texts but struggle with incorporating secondary sources, that they convey a real enthusiasm for the literature they read but are ambivalent about critical methods. (I don't know if any of these things are true. This is purely a hypothetical example.) Actually it's a little more than hypothetical. It reflects broad common assumptions about students and about what is difficult to do in English or more generally in college. These anecdotal things we say about students are largely stable over the years, but interestingly they have little impact on curriculum. We might exchange lore about how we try to address these concerns, but those exchanges do not add up to a substantive change. However we try, to whatever extent we try, anecdotal sharing doesn't create knowledge objects with the force necessary to make change happen.
I think this is clearly evident.
However, we could start an assessment from the anecdotal hypothesis that students struggle with incorporating secondary sources into their writing. We could create a tool that measures these incorporations so that we might construct some knowledge across the program about student performance that might help us fine tune our anecdotal observations and link them together with greater strength. Then (the big step), we might move the issue out of the student and into the network. That is, rather than identifying poor research practices as a student deficiency, we could understand them as a network effect. We could ask, how could we alter the conditions of the classroom and the curriculum to alter this network effect? This goes far beyond the advice of lore because it demands a significant shift in the conditions in the department: a shift that lore is not strong enough to produce.
This is why, when it comes to assessment, I always ask "What kind of knowledge would we require in order to make a substantive change?" That question asks not only about the specific knowledge statement but the process by which the knowledge is constructed. Anecdotes are not strong enough. And my concern for the humanities is that it doesn't believe that any knowledge is strong enough to make such decisions. This, of course, does not mean that curriculum doesn't happen or that changes don't occur. It simply means that we deny ourselves the opportunity to produce knowledge that is strong enough to inform decision-making. Instead we are left with individual feelings, opinions, and beliefs and whatever they amount to. A skeptic might say that this is all that humanistic knowledge has ever been.
But I can't believe that. I can't afford to believe that. If we believe that as humanists we cannot produce knowledge of real value with the strength to make changes in the world, then what would we be doing as teachers or scholars? We would be engaged in some kind of self-pleasuring activity, perhaps with the idea that our performances might instill in others (through some quasi-magical, sympathetic incantation) a similar practice of finding self-pleasure (or aesthetic appreciation) through a purely subjective/cultural/discursive encounter with the objects we study. No doubt there is a strong strand of such thinking in the humanities, especially in English, that goes back at least to Matthew Arnold (though in his case the self-pleasure was imbuded with a chaste religiosity rather than the psycho-sexual implications one probably sees here). However, no one would imagine self-pleasure as the sole goal of humanistic study. We must be able to produce knowledge that has the strength to make changes. And that requires an understanding of how knowledge is constructed and operates in a world that isn't divided into natural, social, and discursive realms. And this is as true for our research and teaching as it is for assessment.