Religious history and big data

Micawberism: ‘The improvident state or habitually optimistic point of view’ (Merriam-Webster definition)

Utter the word ‘big data’ nowadays and research money gushes from the stingy pockets of governments, corporations, and research institutions. Add the term ‘religion’ and we have the equivalent of the Drake Well in Pennsylvania. In this case, USD 30 million has been earmarked – a whopping amount!

This latest offering is for building the world’s first comprehensive online quantitative encyclopaedia of religious cultural history, for which a video clip has also been prepared.

(The video clip is far more ambitious than the text, which mainly focuses on the assemblage of all information on ‘religious cultural history’. In the clip, one goes for the whole hog of both ‘big data’ and the ‘humanities’.¹)

Lured by the fatamorgana of ‘big data’, the logical steps seem to be as follows:

1. Assemble and structure all the available information about ‘religious cultural history’ as well as its pertaining context²
2. Embed the information into computer-legible support (writings, images, sounds)
3. Quantify such information in some meaningful way
4. Parse the information for hidden information

The ambition is breathtaking and worthy of the grant size. The sneaky question is: does it make sense? I shall discuss each step in turn.

Can we assemble and structure all the available information about religious cultural history?

The ‘easy’ part is to create a meta-index covering all the ‘religious cultural history’ that has been published in the West. ‘Easy’ because we may be able to aggregate what is already in dedicated sites.

Difficulties emerge at once, however:

Recording information about their own and foreign religions has long been a Western habit and prerogative. The result was not always faithful to the original. A good example is ‘Hinduism’. Hinduism is a unified worldview British scholars invented for their own purposes (as well as administrative purposes). Unfortunately, it is but a vague – and highly distorted – image of the immense scope of a multifarious Hindu tradition spanning the whole continent and beyond (see The Hindus: An Alternative History by Wendy Doniger). In another vein, our understanding of the many traditional African religious experiences is only now emerging from the shackles of colonialism (see Kongo: Power and Majesty by Alisa Lagamma). What about Orientalism (see Orientalism by Edward W. Said)? Can we justify the use of ‘hear-say’ religion as set out in the West?

In many faiths, there are no Platonic hierarchies of truths leading by stages to a universal: there are only contextual certainties.³ What value or representativeness is one to give to such particular and distinctive guides to the perplexed? The personal and cultural context can become so dominant that swaying between religious worldviews becomes the norm. In Japan, ‘one marries as a Shinto and dies as a Buddhist’. In many countries, diverse religions coexisted or blended. Syncretism was the order of the day (see the Christianisation of Iceland; lore has it that Icelandic seamen in a storm would pray to Jesus and quietly to their old deities). How do we find a usable order?

Floating boundaries exist between mere instruments for internalising social control and immanent or even transcendent religion as understood in the West. In the Kongo, Mangaaka (power figures) secured sanctity of contracts. These fearsome statues were riddled with nails, each nail representing an agreement under the protection of the totem. Violations led the power figure to spirited redress. Is such ‘spiritual’ belief already a religious experience? In fact, even in the West, much of ‘religious experience’ is very practical: help the sick and lame, detect crime, decide on the right time for action (or inaction). Is it religious? Is it an attempt to harness the complex forces of our brain?

Putting it in general terms: what qualifies as ‘religion’? Do we include beliefs and rituals that just aim at living well, but are not concerned with the concluding afterlife, i.e. eschatology?

Can we embed the information into computer-legible support (writings, images, sounds)?

The project aims to deal with recorded or recordable ‘religious cultural history’. The implicit (oh so Western) assumption is that a written record yields an accurate, full, and faithful representation of religious experience. Over generations, furthermore, there is an unbroken and authentic linear succession: Vishnu of yore is today’s Vishnu.

These are doubtful assumptions indeed, though dear to historians of ideas. Culture is not granular information transmitted unchanged across a set of people or down the ages like DNA. With culture and especially ideas, reception, not the text, is the key. This insight is especially true with faith, which is untested against context – it is pure belief. In fact, each person’s understanding of her/his ‘religion’ varies with the personal and cultural context. One cannot put the infinite scope of religious tradition into a few select written straight jackets, nor can these texts be considered in any way representative of lived experience.

Furthermore, the religious experience may be enshrined in the visual, the sensual, and the ritual. Location and sounds, smells, and lights may be essential. The lived experience of religiosity simply defies computerisation.

The West’s obsession with ideas, finally, leaves unexplored the religious understanding and behaviour of the faithful. We have sedulously recorded the utterings and preachings of the shepherd and failed to see the flock’s comportment. Knowing all the details of learned religious discussion says very little about how the faithful live their religion. Furthermore, while we may have some idea about current behaviour, the past is gone and beyond reach. While we may have self-declarations about current religious belief, such utterances are excessively crude.⁴

Can we quantify religious cultural history?

Quantification has driven the West since the turn of the first millennium (see The Measure of Reality: Quantification and Western Society, 1250-1600 by Alfred W. Crosby). We have quantified the material world. Can we make the jump across the big divide to ideas? I am puzzled, in fact, perplexed. Ideas are pure quality. Can we assign a unit of measurement to Christianity or Buddhism? Of course, in the coarsest sense, we can. The consequent loss of information, however, is catastrophic. How do we distinguish in quantitative terms the different strands of belief?

Social facts, and ‘religious cultural history’ is a social fact, do not have objective and countable attributes like length, duration, weight, or field force. Since social facts are facts by ‘common consent’, alternative social constructs can occupy the same conceptual space (see The Construction of Social Reality by John R. Searle).

Can we parse such data for (hidden) information?

As the researchers go about gathering the information, however, they want to quantify religious cultural history in order to subject it to ‘big data analytics’, or so the video clip says. I only have the slightest inkling of what ‘big data analytics’ is. Here is something I have culled from the internet:

‘The primary goal of big data analytics is to help companies make more informed business decisions by enabling data scientists, predictive modellers, and other analytics professionals to analyse large volumes of transaction data, as well as other forms of data that may be untapped by conventional business intelligence programmes. That could include Web server logs and Internet clickstream data, social media content and social network activity reports, text from customer emails and survey responses, mobile-phone call detail records, and machine data captured by sensors connected to the Internet of Things’ (TechTarget).

And sure enough, the narrator of the video clip draws a population in which people act ever so slightly differently and avers that there is more information in the community than meets the eye. This view may indeed be true. Two examples:

1. One can establish a correlation between day and month of birth and big-league performance of Canadian hockey players.
2. Pronouns, articles, prepositions, and a handful of other ‘function words’ reveal parts of the speaker’s personality, thinking style, emotional state, and connections with others (see The Secret Life of Pronouns: What Our Words Say About Us by James W. Pennebaker).

Examples abound as we churn that data, not unlike the Hindu gods, churned the ocean of milk to create reality.⁶ The key element, however, is that all these are transactional data about people who unknowingly leave some sort of ‘cultural footprint’ behind. This has nothing to do with mapping ‘religious cultural history’. Also, in data mining one does not look for intent. One merely observes behaviour or revealed preferences. As a corollary, quantification does not register abstinence, which is a problem when much of the Ten Commandments is couched in ‘Don’t language’ (it is not possible to regress from my avoiding murder to my religious beliefs – if any).

Can a religious experience be quantified in any meaningful way? One can, for example, quantify attendance – an imperfect proxy for belief – but how does one quantify ideas, nuances, distinctions, and differentiations? Quantifications, also, are crucially time- and context-dependent. Time series (as in religious traditions) are poor candidates for quantification.

A gentle caveat

We live in an age of buoyantly emergent techniques. Just look at how DNA sequencing has transformed paleo-history. History has profited from quantification: sweeping generalisations have been shown to hide sloppy research. Quantification can highlight important if neglected aspects of our past – the study of demographics, economics, trade, and inequality enriches the perspective (see Secular Cycles by Peter Turchin and Sergey A. Nefedov). It can give voice to social groups that were not able to record their historical experience. Quantification, however, must be in the service of the particular and the detail: it must flesh out, not flatten out, reality. Quantification should not be Procrustes’ bed.

The Holy Grail, of course, is the Western dream of prediction – the belief that the future is history in the making. Culture-based disciplines easily get the itch to jump on the bandwagon and apply techniques in the Micawberian hope of ‘finding the unexpected’ and prophesying henceforth. Political science and diplomacy are not immune to the lure of quantification and theory-building. I expect them to follow soon – and fall into analogous traps.

_________

1. ‘Our ultimate goal is to include in the database information about all societies known to historians and archaeologists, but such an ambitious goal will, of course, require many years, and will never be truly completed. For the moment, we are focusing on Religious Traditions, but are ultimately interested in data on Polities and Natural Geographic Regions.’
2. ‘Once the database is complete, a researcher can do a targeted search and know where to go for a complete religious and cultural history of any particular topic. We’ve never been able to do this before.’
3. How doubtful this is can be seen by an exercise Stanislavski gave young actors: ‘Express the phrase: Bring me a cup of tea in 40 different ways’ (Esiste un Modo Indiano di Pensare? by A. K. Ramanujan).
4. Over one billion have been baptised into the Catholic faith. ‘Practising Catholics’ are a fraction. Are they all in good standing? No one knows, and in any case: who is to judge? There are major contextual and cultural differences as to who qualifies as ‘practising Catholic.’
5. See Outliers: The Story of Success by Malcolm Gladwell. The correlation rests on the fact that in the annual cohort of players, those who were born first have an advantage over the latecomers. Early success is the basis of later achievement.
6. ‘Samudra Manthan’ or ‘The Churning of the Ocean of Milk’ is one of the best known episodes in Hindu mythology. The story appears in the Bhagavata Purana, the Mahabharata and the Vishnu Purana, and explains the origin of amrita, the drink of immortality. The Samudra Manthan process released a number of things from the Ocean of Milk. One was the lethal poison known as Halahala, which in some versions of the story, escaped from the mouth of the serpent king as the demons and gods churned. This terrified the gods and demons because the poison was so powerful that it could destroy all of creation. Then the gods approached Shiva for protection. Shiva consumed the poison in an act to protect the universe, and his wife Parvati pressed her hand on Shiva’s throat to save the universe. As a result, Shiva’s neck turned blue. For this reason, Lord Shiva is also called Neelakanta (the blue-throated one in Sanskrit).

This post was first published on DeepDip.