The REL Journal Group: Durkheim and Data Edition

The following exchange between Prof. Mike Altman and Sarah Griswold, a student in our MA program, reflects on the recent meeting of the journal reading group, part of our Religion in Culture MA.

Mike Altman: Sarah, for our first journal reading group you chose the article “Durkheim with Data: The Databse of Religious History” from a recent issue of JAAR. What’s the gist of the article and why did you think we should read it in our group of MA students and faculty?

Sarah Griswold: The article is basically an introduction (and justification) for the Database of Religious History. This database is meant to serve two purposes: to be a database for “religious groups” in the premodern world and to provide evidence for a theory of religious evolution. In effect, the database tries to play both fields of holding and providing both quantitative and qualitative data. The article mostly reads as an attempt to draw more scholars in in order to add data to the database.

As far as why I thought we should read it, there were a few reasons. First, as someone with a background in both the humanities and math, I think understanding how and why qualitative data is quantified is really important to understanding and critiquing the purpose and use of databases like this one. Second, as the humanities (and particularly religious studies) moves more and more towards digital projects, we need to be aware of what’s out there so we can emulate what is done well and improve on what is lacking. Finally, the article also offers us insight into the theoretical workings of the project itself. Although titled “Durkheim with Data,” it seemed as though the creators of this project have not critically considered or defined the very categories they have opted to work within, making the move from qualitative to quantitative data suspect. That, I think, can be quite telling of the ultimate success or failure of a project of this size.

MA: As a student in this new MA program that has an emphasis on digital and public humanities what can you learn from this article and what can we as a program learn?

SG: Personally, this article reinforced the importance of thinking through the categories you use when quantifying data. It can be easy to point to something you “know” is religion and label it as such without thinking about why you’ve decided on that label in the first place. It’s also interesting to think about the collaboration across disciplines that these projects require. It would be impossible for one or two scholars to gain all the skills needed to make these things even work. It turns out that group projects exist in real life too and not just in school.

As a program, I think the biggest take away is to pay attention to the developments of these projects. Because the DRH has a capacity to refine their methods, I don’t think they should be entirely dismissed as uncritical. There are positive and negative take aways from critically examining any digital project. Learning more about digital projects and examining their goals and functions can and will tell us a lot about how to move forward in our own individual and collaborative projects.

On Beginnings: Part 18

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

“The signal is the truth. The noise is what distracts us from the truth.”  The Signal and the Noise, p.17.

I wonder how the truth would appear when we found it, and who would validate it once and for all?  Should we assume collections of data “contain” essential “truths” about our own behavior that we can use to reliably predict our future choices?  Or do such statistical and probabilistic applications just plow little furrows where we nurture some irreproachable Mr. Potato Head of signification, permitting us to place and re-place any provisional features?  If so, it might benefit us to be mindful of the selections we make when, for instance, we cite probabilities to plug the gap between indecision and action.  If we place too much trust in what we perceive to be the predictive power of statistical inferences, we may be nonplussed when they inevitably fail to provide exactly the information we think we need.  Then again, there are plenty of times when we conveniently ignore information, statistical or otherwise, when it is at cross purposes to our agenda.  We may allow statistical inference to drive our behavior in ways not otherwise rational; or poo-poo marginal conclusions based on statistical surveys that turned out to have some merit, after all.

Data analyses may also lead us to make inferences not substantiated by the numbers, as in this little gem, describing another PayScale study:

“All these caveats are true. But overall, the PayScale study surely overstates the financial value of a college education. It does not compare graduates’ earnings to what they would have earned, had they skipped college. (That number is unknowable.) It compares their earnings to those of people who did not go to college—many of whom did not go because they were not clever enough to get in. Thus, some of the premium that graduates earn simply reflects the fact that they are, on average, more intelligent than non-graduates.”

I would be interested to see the data that rendered concrete notions of what constitutes intelligence, who has it, who doesn’t, what varieties they possess, and in what measure.  I suspect we can’t easily determine such things.  Or rather, we all can, and freely do.

Part 19 coming tomorrow morning…

On Beginnings: Part 17

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

“We can perhaps never know the truth with 100% certainty, but making correct predictions is the way to tell if we’re getting closer.”  The Signal and the Noise, p. 255.

You can probably guess what I’m going to ask next:  What is randomness?  Is it the quirky and irreducible unpredictability of nature, at the level of the smallest particle?  Or is randomness a temporary inconvenience, an illusion of uncertainty that is really just a lamentable lack of data?  It may seem like a tedious argument, or a matter of semantics, or completely irrelevant to the topic at hand – but our reception of statistical data has a lot to do with how we define and describe randomness, and how much agency we give it in our wider senses of the universe.  If you believe in a deterministic connect-the-dot universe, what seems like chance is rather an unfortunate consequence of the dots we don’t yet see, but can build a better means of detecting.  If you believe in stochastic processes as a fundamental concept of natural phenomena, we can’t say for sure what the hell the dots are doing, or if they can be said to “be” there at all. Continue reading

On Beginnings: Part 16

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

In 1814 the mathematician and astronomer Pierre-Simon Laplace published a now-famed postulate:

“”We may regard the present state of the universe as the effect of its past and the cause of its future. An intellect which at any given moment knew all of the forces that animate nature and the mutual positions of the beings that compose it, if this intellect were vast enough to submit the data to analysis, could condense into a single formula the movement of the greatest bodies of the universe and that of the lightest atom; for such an intellect nothing could be uncertain and the future just like the past would be present before its eyes.”

Complex systems have historically defied attempts at deterministic approaches, not least because the feedback concept of a dynamic system introduces chance, even in an environment in which other variables may be controlled.  “Laplace’s demon,” then, is considered an outmoded approach to physics, but the philosophy of determinism can still be seen in the data-gathering approach of the Digital Age.  The goal needn’t be a harmonious unified theory.  It is enough to take the idea of a lossless network of causation to move along to the next question, heuristically, Bayesian-like, keeping what works and discarding what doesn’t.  Continue reading

On Beginnings: Part 15

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

Maybe another Bill James can offer a non-definitive answer, allowing us to move on to the next question:

“When one views the world with no definite theological bias one way or the other, one sees that order and disorder, as we now recognize them, are purely human inventions.  We are interested in certain types of arrangement, useful, aesthetic, or moral, – so interested that whenever we find them realized, the fact emphatically rivets our attention.  The result is that we work over the contents of the world selectively.  It is overflowing with disorderly arrangements from our point of view, but order is the only thing we care for and look at, and by choosing, one can always find some sort of orderly arrangement in the midst of any chaos…Our dealings with Nature are just like this.  She is a vast plenum in which our attention draws capricious lines in innumerable directions.  We count and name whatever lies upon the special lines we trace, whilst the other things and the untraced lines are neither named nor counted.”  Continue reading

On Beginnings: Part 14

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

“Make it seem inevitable,” Louis Pasteur advised his students preparing to publish their research, in the oft-cited apocryphal chestnut.  When we present statistical data as though the data itself harbored some perfect implicit revelation, we are doing just that.  When the data “misleads” us, we are doing that yet again.  Even the polling data Nate Silver relies on is subject to our vacillations between obstinate fealty and obstinate skepticism.  There are times, of course, when polls really do get it wrong, but it doesn’t seem to affect their credibility until the results clash with our agenda.  Or when elections that don’t turn out like we want can be deemed “flawed.”  We laud the numbers when it suits our purposes, then call compilations of those numbers tainted when they produce outcomes we consider undesirable.  Is the data “bad?”  Did we collect it imperfectly, or imperfectly interpret perfectly true information?  Are we wishy-washy?  Or is this just how we shimmy through life, alternately contesting and consenting in the service of our momentary aims?  Do we hold static views in a mutable world?  If we did, we wouldn’t have to take polls so compulsively – but we’re fickle.  We’re duplicitous.  We strategize.  Even with a constant showing of hands, a constant checking-in, political polls aren’t a reliable indication of an election outcome. Continue reading

On Beginnings: Part 10

This essay (serialized here across 24 separate posts) uses words and numbers to discuss the uses of words and numbers — particularly examining evaluations of university degrees that employ statistical data to substantiate competing claims. Statistical analyses are crudely introduced as the mode du jour of popular logic, but any ratiocinative technique could likely be inserted in this re-fillable space and applied to create and defend categories of meaning with or without quantitative support. Questions posed across the series include: Is the data informing or affirming what we believe? What are the implications of granting this approach broader authority? The author, Melanie Williams, graduated from UA in 2006, with a B.A. in Anthropology and Religious Studies.

 

Representations of “truths” bolstered by statistical data need not be deliberately misleading to be subject to tendencies B.A. holder Darrel Huff called “statisticulation.”  Any data we might collect is a product of the question we ask, and the ways in which we seek to answer that question.  Of course we are “manipulating” the numbers – we can’t help doing so in the process of gathering bits we deem relevant into the little pile we use to build up notions of order, character, and causality.  Is there a way around this dilemma?  Depends on whom you ask:  Modern, silicon-based engines of technology, some contend, have rendered obsolete the traditional vetting cycle of facts – observe, hypothesize/model, test – by providing more precise methods of measurement and a vast interconnected repository for storing and sharing data.  Data no longer needs a model, in this approach – it can be collected and analyzed statistically by high-speed processors.  The “End of Theory,” as Chris Anderson opined in 2008, may be nigh, allowing scientists to kick back with a margarita while pattern-detecting algorithms process data that will autogenerate yet-unknown points of inquiry – no never-adequate scientific model required, in all its messy human-ness.  You could call this Big Data method progress.  You could call it a pendulum swing.  Or you could call it a waste of time to contrast to classical methods, since no one actually believes data is worth collecting for the sake of collection alone, with no defined mechanism for selecting and interpreting it.  Who would store all that?  Besides the NSA? Continue reading

Caution: Technical Terminology Ahead

Picture 7I see posts like this on social media all the time (click it if you’re dying to find out what those 22 words are); what I think drives them is a general failure on the part of many to understand language as a tool used by groups to achieve a variety of social ends and not a universal medium in which we all just naturally swim. For only if we assume the latter would we be shocked to find out that what we mean by some word is not what they mean by it. Continue reading

Finding or Fabricating?

pyeMichael Pye, the onetime General Secretary of the International Association for the History of Religions (IAHR), and respected specialist in the study of Japanese religions, recently presented a keynote lecture — “Digging for Theory” — at a conference at the University of Göttingen, Germany. Continue reading