My friend and colleague, David Miller (English, South Carolina), sends this along:

The University of South Carolina is seeking to appoint a Research Assistant Professor in Digital Humanities.  Here is the text of the job advertisement:

Research Assistant Professor, Digital Humanities

The University of South Carolina College of Arts & Sciences seeks to appoint a specialist in Digital Humanities to support collaborative research and develop infrastructure for a proposed Center for Digital Humanities at South Carolina.  Candidates should have a PhD in a humanities discipline, along with a strong background in computational research methodologies.  Important skills for the position include an imaginative approach to collaborative research, efficient project management, strong grant-writing abilities, the ability to build coalitions and work cooperatively with a variety of partners, and the ability to communicate effectively with specialists in a variety of disciplines including Computer Science.

Nine-month salary $50,000 with benefits; summer salary to be supported by funded projects.  Based on performance the position is annually renewable.  Send letter and resume by February 15, 2010 to David Lee Miller, Director, Digital Humanities Initiative, Department of English, University of South Carolina, Columbia SC 29208.

The University of South Carolina is an affirmative action/equal opportunity employer.  Minorities and women are especially encouraged to apply.  The University of South Carolina does not discriminate in educational or employment opportunities or decisions for qualified persons on the basis of race, color, religion, sex, national origin, age, disability, sexual orientation or veteran status.

Tweet This Post

“We must jump off Lévi-Strauss ‘s bus one stop before he does.“
Wendy Doniger, The Implied Spider, p. 149.

It is hard to overestimate the infliuence of Claude Lévi-Strauss, a French philosopher turned anthropologist who liked to be known as a craftsman, on the discourse of the humanities and social sciences in the second half of the twentieth century. Operating among a surplus of great thinkers (especially in France) during a period of the devaluation of the role of intellectual, his name nevertheless stands out not simply for having founded one of the most important intellectual movements of the century — structuralism — but for having defined the agenda, along with Wittgenstein, for a host of major thinkers up to the present time.  Without Levi-Strauss, there is no Foucault, no Derrida, no Bourdieu, no Lyotard to both absorb and react to the idea of structure and the great anthropological tradition which served to flesh out the idea.  For without Levi-Strauss, the vast storehouse of descriptive ethnography that he discovered and fully absorbed in America, in the New York Public Library, while in exile during the war, would never have reached Paris.  The significance of these materials — many of them dry catalogs of cultural traits interspersed with tentative observations about the nature of culture — would never have caught the attention of the European eye.  And without his role as cultural vector between American empirical ethnography and French philosophical anthropology, the deep well of ethnography in general might never have informed the imagination of deconstruction, poststructuralism, and postmodernism.  Evidence for the influence of this source can be found everywhere, both implicitly, as in Foucault’s appropriation of the idea of mana to theorize power, and explicitly, as in Lyotard’s strong reference to Lévi-Strauss and the ethnographic example of the Cashinahua in the Postmodern Condition.  Surprised by his fame and notoriously absent from the activism of the 1960s, Lévi-Strauss was the unmoved mover for an entire system of thinkers whose lights shone the brighter for their proximity to him.

Today the bus has stopped.  For those of us who have followed Doniger’s advice — and you have to — perhaps now we may consider getting back on for a few more stops.

Tweet This Post

Musings on Zacks, the Hippocampus, and Kafka

NOTE: This article has been modified slightly since I originally posted it.  The section on the hippocampus has been subordinated to an inline note.  It is interesting (at least I think so) but not part of the main argument I’m trying to make here.

A few weeks ago I had the good fortune of seeing Jeff Zacks give a talk entitled “Film, Narration, and Cognitive Neuroscience” at the Curry School of Education at U.Va.  He was invited to speak by my colleague Glen Bull as part of the school’s Tea & Technology series (every Thursday at noon in Ruffner Library if you’re around).  It was an excellent talk not only for its content, but for Zacks’ gift of clarity.   At no point did the science overwhelm, and the talk had the useful effect of raising lots of interesting questions in
my mind and in the minds of others.

Zacks’ thesis is relatively straightforward: human beings have the capacity to perceive events as a series of discrete segments, and this capacity to segment events is correlated with the long-term memory of those events, as well as the ability to focus on and form concepts about the goings on in an event.  There are big cognitive pay-offs to performing this work of segmentation, pay-offs which  (predictably) are thought to correlate to evolutionary advantage.   Interestingly, the boundaries that people spontaneously define for events — such as a sequence in a French film or a film of a man washing dishes (two examples used in Zacks’ experiments) — are widely shared.  Where there is variance, subjects are either older or suffering from dementia (which gives me a really good feeling about the aging process).  The mechanism for segmenting thus appears to be hard-coded and deeply rooted, probably in the hippocampus.

One of the most interesting parts of Zacks’ thesis is the process by which event boundaries are perceived, or rather, defined by the brain.  Apparently, boundaries are inserted where the brain experiences what Zacks calls “prediction error” — when things break a pattern of repetition and thus signal to the brain a boundary that is used to construct the temporal model for the event — its typical sequence.   Zacks did not mention this, but I would be surprised if this capacity were not in some way connected with the capacity of the brain to perceive information in Shannons’s sense.  For prediction error is an excellent name for the newness that his negative entropy equation defines.  My tentative hypothesis is that we must be wired to both perceive information in this sense, and to make use of it in the formation of perceptual structures which, in turn, become the materials from which cognitive structures are built.

Now I mention this because the response of the audience — comprised mainly of educational experts — and of Zacks himself is that one practical lesson from his research is that creators of narrative content, such as film, should make an effort to provide more obvious segmentation in their products.  Clearly, if this is how the brain works, we should work this way too.

I think this is a major fallacy that pervades the reception of brain science research.  People tend to assume that if the brain works a certain way, then so should we. I call it the fallacy of brain-behavior mirroring (at least until I come up with a better name), a more recent variant of the mirror of nature fallacy described by Rorty.  For example, since we know (since Kant and then Gestalt psychology) that the brain actively constructs objects from experience, then, the argument goes, we should have students actively construct things too.    But clearly, if the brain already works a certain way, then it works that way — in spite of, or perhaps because of, how we behave, oblivious to a detailed description of its workings.  In the case of event segmentation, it seems clear to me that Zacks’ research suggests an opposite lesson — it says that people who can perceive prediction failures can also segment experiences, and thereby gain cognitively.   If so, then the lesson is to get good at perceiving and creating event boundaries, which requires not pre-segmented media, but the opposite — hard to grasp art, stuff that violates expectations and rewards the perciever with a different perspective.   In fact, giving students media with well defined boundaries may cause their capacity to construct boundaries to atrophy, much as caffeine causes our adrenal glands to shrink. (I know, it’s a good reason to stop drinking coffee.)

Now it turns out that some research form the University of California at Santa  Barbara corroborates this view.  According to the headline of the New York Times,’ piece in this work, nonsense — such as the absurdist work of Kafka — sharpens the intellect.   (Another summary of this research is entitled This is Your Brain on Kafka.)    Essentially, your brain has to work hard to make sense of things, which it is evolved to do, and this has the side effect of making you smarter, or at least sharper in the period following the effort to make sense where there is none.  And what is nonsense but prediction failure on a large scale?

So, what is the pedagogical and media design lesson here?  Learning Teaching is not about making content easy to ingest, it’s about creating environments where students can play this game of meaning formation, which isn’t always stress-free.  Marketers may disagree, but they are in the business of indoctrination, not teaching.

Tweet This Post

When I get the time, I’m going to write a vocabulary creation language to support structuralist text interpretation.  It will consist of two specs: one to handle the marking up the surface features of text, such as rhetorical figures and tropes.  This will be based on my work with the Princeton Charrette Project and it will likely incorporate some ideas from Steven Bird’s work on annotation graphs.  The second will be either an extension of or a variant of SKOS and|or OWL designed to represent extracted symbolic structures.  It will incorporate predicates to handle relations of signification, such as has_part, has_analogy, and has_metonym, between the elements represented in the first language.   At a larger level, I want to represent holistic dimensions such as context and level, as well as narratological things like encompassment, transformation, inversion, and liminality.

One of the big problems I see in this project is an apparent limitation in RDF to support triples about triples.  For example, an analogy is a relation between structures, not terms.  The assertion A : B :: C : D is, at minumum, an assertation about the relationship between two assertations, A : B and C : D.  (The predicate of the assertions themselves is usually X has_part Y.)  An anology looks something like this then:

[A has_part B] just_as [C has_part D]

The easiest way to accomplish this task would be to provide URIs for each RDF triple.  I haven’t seen a general solution to this problem.  I know I can create local URIs within a specific triple store, and use these in triples.  But I need to define an RDF triple as a datatype first.  And I anticipate problems further downstream; I wonder if the current RDF toolset is designed to handle indexing and inferencing of these kinds of triples.

If anyone has suggestions about how to handle this issue, I’d be glad to hear them.

AFTERTHOUGHT:

After writing this, it strikes me that to say that two triples are analogous is just to say that they share a predicate–so long as that predicate is sufficiently specified.  To assert an analogy, then, is to assert that such an identity is important or relevant in a certain context.

Tweet This Post

Comparative ontology asserts that humans already have ontologies, and that machine ontologies are both projections of human ontologies (those of the numerati) and material agents that intervene in the ongoing reproduction of ontologies (everyone else’s).  Developers of ontologies for the web of linked data would do well to understand the nature of human ontologies, as well as they way machine ontologies intervene in the ongoing construction of social life.

Human ontologies are not like ROM programs, hard-wired into our brains and executed without modification; they are designed to be reprogrammed through engagement with the world. They are one of our most effective adaptive traits.

Ontologies are adaptive

Anthropologists have studied ontologies in the wild for a long time, under the various categories of “structure,” “symbolism,” “culture” and “collective representations.”  One of the most important contributors to the study of ontology is the American cultural anthropologist Marshall Sahlins.

Sahlins’ principal theoretical contibution to cultural anthropology has been to retrieve the concept of cultural structure of the ahistorical, formalist, and mechanistic conception developed by Lévi-Strauss, whose own work on mythology belies his more theoretical pronouncements.  Rather than separating structure from event (and history), and locating the former deeply within a universal mind–like a camshaft responsible for the jigsaw puzzle of culture–Sahlins focuses on what he calls the “structure of the conjuncture” of structure and event.  History emerges as a culturally distinctive second-order structure that results from the ongoing work of categories in praxis. So categories have a structure, but that structure undergoes reevaluation and change as it is applied to the world.

In Islands of History Sahlins describes the process of cultural (ontological) change in terms of the “risk of reference”: as cultures classify things in the world–as they deploy ontologies–they also put these ontologies at risk.  For things in the world do not always behave as classified, or planned.  Even the sun  has an occasional eclipse.  Although the keepers of culture–from priests to grandmothers–try to enforce adherence to the categories, the behavior of things will inevitably contradict the categories and call for their revision.  Sahlins reads the Hawai’ian’s classification of Captain Cook as Lono as just such a world changing event.

Ritual is one mechanism humans use to synchronize the world with world view.  As people grow, for example, and change statuses, rites of passage are used to mediate this “contradiction” and reclassify people so that they can fit into the system.  Another mechanism is prophecy, where  the reverse is true–world views are aligned with a world that has changed.  Millenarian movements are the classic example of this: a prophet emerges who can make sense of the new in terms of the old, but changes the old in the process.

Rituals and prophetic movements are the original forms of change management.

This is the ongoing work of culture.  Cultural reproduction is never mechanical.  That is one reason we humans have history.  There is always a disproportion between words and things, plans and situations.

Databases (and the point of this post)

Now, a data model, such as a set of tables and fields in a relational database, an XML schema of elements and attributes, or an RDF vocabulary of classes and properties, is a plan, a schema of classification.   And database applications, like rituals and texts, have their own forms of referential risk to contend with. They classify the world and, in the process, both effect the world they classify and open themselves up for revision by that world as it changes.

For example, the categories produced by a requirements elicitation process for an application designed to improve some workflow, and encoded in a database that sits at the bottom of an application stack, may not accurately represent the workflow as it is actually practiced, and as it will inevitably change as new developments take place–changing personnel, clients, strategic plans, etc. The database, then, is put into a situation–the situation of the conjuncture–into which its categories are at risk.

In this situation, databases are like texts–they are built on the armature of a hard-coded ontology, and they can move beyond their original domain of applicatibility.

But unlike most texts, and very much like sacred texts, database applications (and their administrators) are usually given a central position within an organization.  They are often deployed as key elements of an enterprise architecture that calls the institutional shots.  Thus they can insulate themselves from referential risk.  They can force conformity to their logic–as Michael Wesch’s New Guinea villagers redesigned their settlement pattern to conform to the government census–or they can produce a black market of behaviors in an organization that bypasses the database governed workflow.  This is what faculty do who are forced to use an LMS but would rather use Google Docs.

Comparative ontology can help here.  If we view ontologies as always situated, then we should (1) design systems for maximum flexibility and adaptabilty, and (2) learn a lesson from the ritual life of peoples around the world and throughout history: engage our ontologies in constant reevaluation and modification, making the world (of our organizations) fit where appropriate, and also refining the categories to fit the world.

To meet the first challenge, we shouldn’t create overwrought ontologies, but rather focus on just enough classification to achieve the effects we need. Usually, the effects we are most concerned with are connecting people to people, people to information, and information to information, in as few links as possible.

To meet the second challenge, we may want to refine what we mean by “social operating system”–for that is precisely what a ritual system is.  Maybe it’s time to follow McLuhan’s advice and exploit the ritual effects of the electric, in order to mitigate and shape the more dangerous effects of the electronic.   When we build ontologies, maybe we should also be thinking of the physical and virtual spaces in which they will be deployed, and the material and digital artifacts that will be their vehicles of expression.

Tweet This Post

I am all for user-driven design methodologies. My instinct is to distrust the Central IT ethos of “we know better” because “we think more rationally about things” and all that.  That perspective is based on a simultaneous over-valuing of a linear, rational notion of process (“planning”) and a grudging acceptance of user behavior as “cultural” and therefore outside the scope a requirements gathering process.

The term “non-functional requirements” speaks volumes and captures the Central IT attitude very well.  Under that category, the whole point of effective software design is swept under the rug.  We know that software will be most effective when it adapts to user behavior and vice versa, but we often sidestep that issue, hoping for incremental, evolutionary changes to produce the desired effects over the long run.  We miss the opportunity to innovate, leaving that to the less timid.

But I also find that user-centric methodologies are based on naive assumptions about what users want, or who The User is, or what the point of the user research is in the first place.  Unless you have a very restricted audience for your software–and admittedly one often does–it is very difficult to translate the views of a few people, whether captured by focus group, survey, or even participant-observation, into generalized principles for an application.  Ultimately, good design is what works, and we retrospectively attribute success to our process.  But we really have no clue.

What is it that one is capturing by user-centric research, anyway?  The attitudes and dispositions within a class of individuals?  This can’t be it.  User attitudes and mental models are highly variable, and they are mutable because humans are adaptive, more than we think.  If you build software based on some static notion of what users want, what they say they want, you will miss the effect software has on redefining what they really want.  This is because users inhabit cultural environments, and software inevitably has effects on those environments.  If you focus too much on the abstract user–what’s “in” the user–you will often have the feeling of the goal posts moving.  Or you may end up dismissing the user altogether as fickle and irresponsible, and go on with your own design ideas.   If you design software for a living, I am sure you know what I am talking about.

I think the proper focus of user-centric software design has to be the user-in-context.  That is, not the user but the Situation.  But sutuation defined in a specific, rigorous way.  Situation as the objective, institutional framework of power and infrastructure in which people work.  This is difficult terrain to study, hedged in as it is by all sort of taboos and misrecognitions that keep the social gears moving.  Let me give you an example.

One of the areas where the Central IT software design ethos dominates is in the area of document management.  Two factors drive the design of solutions: (1) developers assume (know) that paperlessness is a Good Thing, and (2) the paper-based workflows that users are enmeshed in are so crufty, complex, and idiosyncratic that it is impossible for users to describe them in enough detail to re-engineer them.  The result is that the digital document management solution will almost always build around people’s behavior, or else it will break workflows where it has to.  So, instead of stepping back and rethinking what the data flows entailed by a paper form entail, or taking advantage of the metasocial moment and asking Why are We Doing This in the First Place, document-logic is reproduced in the software.  The efffect is not to reproduce the old way, and make it more efficient.  It is something unpredictable and bound to have hidden consequences, not all of which can be good.  Most likely, we’ve preserved the notorious stupidity of bureaucracies and have ensured its continued survival in a mutant  and more powerful form.  Because once categories get encoded in institutional databases, the tail wags the dog.  Think health insurance.

So, what to do?  I suggest that we pursue theory-driven design.  We actually try to make sense of the sociology and anthropology of bureaucracies and operationalize the best ideas in these discourses as design principles.  We think of how software behaves as an assemblage of artifacts in a living cultural environment.  This is not social engineering, nor is it to tread the tired path of “organizational behavior,” a field that is too closely tied to the executive perspective.  It is to pursue a rich, empirical understanding of software in the wild, or at least, the office.

Theory-driven design is not anti-empirical.  It is the opposite: for a good theory generates testable hypotheses.  It gives a framework to user-centric research beyond the unanswerable quest for what users really want.  As they say, there is nothing so practical as a good theory.

A good starting point might be to take Ted Nelson’s ideas about documents and hypertext and combine them with, say, David Graeber’s critical anthropology of bureaucracy.    Not to condone the anarchism of Graeber, but to lever the authenticity of perspective he brings to a discussion about the role of documents in the organization.  Reading his essay, “Beyond Power/Knowledge an exploration of the relation of power, ignorance and stupidity,” it is hard not to believe that a radical rethinking of the document, and document-logic, would not benefit from his perspective.

Tweet This Post

I suppose it is the prerogative of different generations to simultaneously dismiss and retrieve old ideas by introducing new words for them. I have in mind words like “metacognition” and “knowledge management.” In both cases there is an existing word that more or less describes the referent of the new(ish) word: epistemology and education respectively. Both metacognition and epistemology refer to, roughly, the activity of “thinking about thinking,” and the core mission of education is the management of knowledge — producing it, storing it, reproducing it, etc.  However, in each case, the intent of the new word is clearly different from the older one, and this difference can be attributed to a different organizational context: knowledge management is about education and research in corporate settings (now defined as “knowledge producers”), as opposed to society or the world at large, while metacognition has flourished within the relatively narrow context of academic departments of education.

But why the complete absence of the old words in the new discourses? Why not call metacognition something like “applied epistemology”? Or knowledge management “corporate education” or “corporate teaching and learning”?   It can’t be for lack of familiarity with the older words.   Nor can we assume that the newer words are more “sticky” and easier to use; that just begs the question.  I think it’s clear that the problem with these constructions is their connotations: they carry too much semantic baggage.

But now here’s the thing: the new words do not simply stand alongside the old ones, they actually seem to take their places.   The new words take the place of the old words at an abstract level–but then replace the implicit social meanings in the process.  The effect is to implicitly usher in newer or different institutions in the space reserved for the old. So knoweldge management is about education, yes, but education in a business setting where knowledge is viewed as a competitive advantage, not a general good for the betterent of humankind.  And, eventually, this will have implications for education itself, as essays like “Applying Corporate Knowledge Management Practices in Higher Education” become more common.

Similarly, metacognition is about epistemology, but not as an abstract philosophical concern, nor one tied to the remote activity of a purely scientific enterprise as it once was; it is epistemology in the service of classroom teaching and learning, where the users of the word no doubt think it belongs. So the effect of the word “metacognition” is to usher out the ivory tower and to replace it with the more populist institution of the classroom. And this meaning is consistent with the current ethos of educational populism, as expressed in wider ideas like connectivism.

So language really does embody the social:  these words are actually the encodings and amplifiers of social changes happening right now.  Perhaps those of us familiar with the older names of things would do well to note these shifts and pay attention to their institutional commitments.

Tweet This Post

Symbols — from core symbols like the Virgin of Guadalupe to abstract ones like the whiteness of Melville’s whale — fix and generate ontological categories.  How and why this happens is a question of deep interest to me, but that it is true seems obvious and well established.  Human beings create symbols like plants produce oxygen, and symbol formation is inextricably bound to a defining trait of human beings — rich, discursive, and always already metacognitive  language (human beings have always talked about talking, a practice that must be regarded as intrinsic to human language).  Linguistics up to now, wedded as it has been to a Chomskyian Cartesianism, has missed this role, although philosophers have not lost sight of it.  As Ricoeur wrote, “the symbol gives rise to thought.”

The relationship between language and symbolism is complex and (still) not well understood.  My own view is close to that of the (admittedly discredited in its original form) generative semantics school, associated with George Lakoff.  I believe that categories and rules are in some way generated by the transduction of meaning that takes place between neural representations of concrete objects.  Discursive language — not “deep grammar” — tries to fix these meanings in propositional form, but the symbolic substrate has a dynamic quality, in no small measure do to its adaptive nature in response to what Merleu-Ponty called the “primacy of perception.”

With writing and then printing, and the monopolization of explicit knowledge, in the form of written records, reference works, etc., by governments, universities, etc., the relationship between discursive fixation and embodied symbols becomes tenuous and contested, resulting in a mind/body problem unfamiliar to ritual societies.

In any case, a number of practical observations follow from this tenet, which I will quickly enumerate, and hopefully take up later:

1. Human ontologies are not plans.
2. Human ontologies are overdetermined.  That is, there is always more than one way to express an ontology  The fixing of meanings will always fail if the goal is to create non-overlapping, non-redundant descriptions.
3. Human ontologies are rhizomic.  In their natural form, ontologies are not hierarchical.  Rather, the hierarchical representation is one form of serialization that works well because of its analogy to kinship (see Durkheim and Mauss, Primitive Classification).
4. Human ontologies are local and situated.
5. Human ontologies evolve.

Tweet This Post

Just as institutions require shared ontologies to function, so are institutions involved in creating the categories that nake up ontologies.  Admittedly, assigning agency to institutions poses a number of questions that need to be answered; I won’t attempt that here.  Essentially, I follow Mary Douglas (see video below), especially her How Institutions Think.  The categories and rules that comprise human ontologies follow and enable a practical logic that in turn enables sustainable human interaction.

A corollary idea to this is that ontologies exist (in large part) to mediate social action.  They are the result of human beings’ mutual calibration of individual cognition through collective interaction.  Typcially, this calibration takes place through ritual.  But media (old and new) — which grow out of but eventually displace ritual — also take on this role. (McLuhan’s frequent reference to ritual to describe the effects of new media is telling.)

“Social Life Makes the Categories“
The late Mary Douglas being interviewed in 2006
about her book Purity and Danger.
See the full interview and films at ScienceStage.

Tweet This Post

This is the first in a series of posts in which I define some of the tenets of comparative ontology, in order try to flesh out its significance to the work of making and using RDF vocabularies.

The overwhelming conclusion to be drawn from the ethnographic record is that human beings are surprisingly structured in their thinking and behavior, even when that behavior seems to be random and non-linear.  Although it is a commonplace to observe that social life is inherently messy, unpredictable, and resistant to capture by physics-like laws (post-Einsteinian included) it remains true that patterns of culture are remarkably widespread and persistent.   Languages, marriage practices, calendar systems, gift exchange systems, markets, etc. — essentially, any functional human institution — all rely on shared categories and rules to operate, and these are discoverable and describable.  The mistake of the structuralists was to conceive of these categories and rules as logical in a formal, almost scholastic sense, like a plan that agents follow strictly.  Instead, it is more likely that they exist as dispositions that constrain behavior and encourage improvization, as Bourdieu describes in his idea of the habitus (which he got from Mauss, by the way).

To the extent that formal RDF ontologies are meant to mediate human-computer interaction (and not simply allow computers to share information), ontologies should be designed to interdigitate with the categories of their human participants.  Machine ontologies should be interoperable with human ontologies.  They should be designed to encourage the symbiotic development and evolution of human collective representations (to use Durkheim’s expression), given the role of the networked computer and computer network as an institution in its own right.

Tweet This Post