The risk of talking about context is the temptation to treat it as some undifferentiated thing. Just add three more cups of context, stir, and voila—a valid interpretation of the facts. I tried to show by example that this was not so, but the distinctions among context are worthy of independent investigation.
I am trying, lately, to be better about making return journeys through texts that I have read already. One such text, which I found very insightful the first time through, is Don Lavoie’s collection Economics and Hermeneutics.
Since reading it the first time, the role of models has become more elusive to me, one of many questions that feel unsettled to me lately after having felt settled for some time. I like, for example, Joseph Heath’s treatment of agency theory as not literally true but still useful. But lately that relationship—of the model to truth, even of a useful model to truth—has become highly ambiguous in my mind.
I remembered Richard M. Ebeling’s chapter of Economics and Hermeneutics, “What is a price? Explanation and understanding (with apologies to Paul Ricoeur),” which I had previously seen cited elsewhere and was what led me to the collection in the first place. I did not really understand it the first time through. So I decided to check out the Ricoeur paper that inspired it, “What is a Text? Explanation and Interpretation,” and to take a return tour of Ebeling’s piece afterwards.
Ricoeur’s piece was a disappointment. He was grasping at an answer to the same question—how can models aid us in our interpretation of a reality which is larger than any model? But the answer he provided did not persuade.
Ricoeur’s title, along with Ebeling’s, are modeled on Wilhelm Dilthey’s classic formulation of explanation, as the mode appropriate to the physical sciences, and understanding, as the mode appropriate to the Geisteswissenschaften, the human sciences. Ricoeur and Ebeling wish to argue that both modes are actually appropriate to the latter.
Ricoeur argues that we must start with explanation, and then complete “the hermeneutic arc” with understanding. In his specific examples, we begin to read a text through the structuralist lens of a Levi-Strauss, and then we move from there to the more subtle art of interpretation.
This is, if anything, backwards. While reading The Illiad or The Great Gatsby, must we begin from the lens of an over-arching theory from Marx, Kant, or Durkheim? If so, then hardly anyone has read anything appropriately. No, we must instead begin with what Gadamer characterizes as an openness to the text, to the questions it poses to us, to its subject matter. Only then, to enrich our understanding or to correct some misunderstanding, or simply to answer some question the text poses to us, do we bring models into the picture. A critic, for instance, may attack our interpretation on the basis of some trope—a kind of model in fiction—that we have utterly misunderstood.
Returning to Ebeling, I found him just as disappointing as Ricoeur, now that I had actually read the latter. As I said, if there is a “hermeneutic arc” at all, openness must come first, not second. And there is in fact no hermeneutic arc, but a spiral of provisional judgments, continually revised.
Deirdre McCloskey’s contribution to the collection, which I found unremarkable at the time (having just recently read her Rhetoric of Economics and Knowledge and Persuasion in Economics) in fact addresses this question more appropriately. From the chapter in question, entitled “Storytelling in Economics“:
There seem to be two ways of understanding things: either by way of a metaphor or by way of a story, through something like a poem or through something like a novel. When a biologist is asked to explain why the moulting glands of a crab are located just as they are he has two possibilities. Either he can call on a model—a metaphor—of rationality inside the crab, explaining that locating them just there will maximize the efficiency of the glands in operation; or he can tell a story, of how crabs with badly located glands will fail to survive. If he is lucky with the modelling he will discover a mathematical model with analytic solutions. If he is lucky with the storytelling he will discover a true history of some maladapted variety of crabs, showing that it is dying out. Metaphors and stories, models and histories, are the two ways of answering ‘why’.
It has doubtless been noticed before that the metaphorical and the narrative explanations answer to each other. Suppose the biologist happens first to offer his metaphor, his hypothetical individual crab moving bits of its body from here to there in search of the optimal location for moulting glands. The listener asks: ‘But why?’ The biologist will answer with a story: he says, ‘The reason why the glands must be located optimally is that if crabs did a poor job of locating their glands they would die off as time passed.’ A story answers a model. Likewise, a model answers a story. If the biologist gives the evolutionary story first, and the listener then asks ‘But why?’, the biologist will answer with a metaphor: ‘The reason why the crabs will die off is that poorly located glands would serve poorly in the emergencies of crabby life.’ The glands would not be optimally located: that’s why.
Metaphor and narrative are two types of context.
Consider the two “bundles of context” I offered in this post, to aid in interpreting this post by David.
The first bundle attempts to explain his post by telling a story about events in his life which led up to the writing of it.
The second bundle attempts to show that the post is meant to be read as a metaphor, for an experience that readers are likely to have had—the experience of never feeling like one has enough knowledge to be on solid ground.
Combining the two bundles, we can see that David’s post is also a critique of poorly constructed metaphors—the metaphor of “knowing enough,” used to bludgeon people who disagree, without actually elucidating when enough is enough. Seeing David’s post as a metaphor gives us the biggest bang for our buck, in terms of undertanding; but embedding it in narrative continues to enrich that understanding.
Joseph Heath’s paper on agency theory, from this perspective, can be seen as arguing that the theory in question is a useful metaphor, but by itself does not suffice. Unsupplemented by other metaphors and narratives, it invariably leads us into error.
The distinction between metaphor and narrative in part is one of contingency. Metaphors are by nature less contingent than narrative. A general model of how hearts function is less contingent than the specific evolutionary narrative of how hearts developed in a given species.
Much more than this can be said, and should. The questions raised here connect with concepts in linguistics, literary theory, metaphysics, and ontology—at minimum. But whatever it might mean to know enough, I certainly do not know enough of the appropriate material from these fields. So I will leave it at that, for now.
3 thoughts on “Torturing the Data, Attending to the Text”
A very interesting take on this whole issue. I found your perspective useful and insightful, if not enough to change my mind, in general. But thank you for taking seriously my very modest contribution to this topic.
Thanks for taking the time to read and leave your gracious comment. Since you are here, would you say that I read Ricoeur and you correctly? Or did I misrepresent your views? (For an amateur like me, that’s a very likely scenario)
Much of this stuff is way over my head, but I recommend you read Kalev Leetaru’s article at Forbes on the Federal government twisted data gathering and analysis practices… http://www.forbes.com/sites/kalevleetaru/2016/03/29/system-failure-what-ive-learned-as-a-data-scientist-in-washington/#5b888a0b4ef4