How do you know?
This article first appeared in the Minden Times on August 23, 2023.
We live in a time of a lot of information and very little certainty. And that makes us uncomfortable, since human beings, by nature, have a pressing need for certainty. We kinda like black and white: grey makes us queasy. The battle to pin down Truth is therefore as old as mankind.
The battle has been fought in many ways. In 1850s Middlemarch, Casaubon, the 50-something pedant, compiles tidbits from Greek and Latin classics to discover the Key to All Mythologies: a bit like a pre-digital ChatGPT. Young Ladislaw, travelling in Europe, knows that the new ecclesiastical truth is being explored through ardent argument in German. Casaubon does not read or speak German, rather like many of my generation are illiterate in techno-talk. Dorothea, the gender-challenged, intellectually-aspiring naif, wishes that this rampant race for Truth, regardless of how it is practiced, will result in something that improves the daily life of mainstream people.
Jill Lepore, in an article in the Apr 3/23 New Yorker entitled Data-Driven, explores the current mud-wrestling for the true source of Truth. She starts with a delicious send-up of what the development of Artificial Intelligence would look like if it were humanized: under-employed university graduates are paid to read books, the books are burned, the brains are pickled and stored in jars on a shelf. She adds a metaphorical imagining of the consequences of the unacknowledged insufficiency of combining those brains: the Truth is there are safe and poisonous fungi but the lesson fails to adequately identify the difference and the teacher takes no responsibility for – is blissfully ignorant of -- the deaths that ensue. Lapore proposed a welcome typology of knowing that goes beyond the current love affair with ‘‘Data’, as if it were the only place you can find any answers, as if only data tells because only data sells.’
There are four kinds of knowing, she says, and visualizes them as organized for accessibility in four drawers entitled ‘Mysteries’, ‘Facts’, ‘Numbers’ and ‘Data’. In the top drawer, closest to Heaven, are Mysteries, things only God knows, ‘like what happens when you’re dead.’ People use theology to discern Truth for the purpose of achieving salvation, whatever that is determined to mean.
In the Facts drawer, which began to fill when Science was discovered, are ‘things humans can prove by way of observation, detection, and experiment’ studied through ‘law, the humanities and the natural sciences.’ I think it is essential to existence that we act as if many things that are unprovable are indeed facts, but this is the core of countless unresolvable arguments between myself and my brother, who is an engineer and professes to believe that people operate in an if-this-then-that linear way, and can’t likely be laid to rest in this paragraph.
The Numbers drawer holds ‘the measurement of anything that can be counted’. Statistics were originally defined as numbers gathered by the state, associated with the rise of public administration, the handmaiden of social sciences. Governments – ah we do know this – use statistics to describe the world they govern and base action on that Truth. (The phrase ‘lies, damned lies, and statistics’ is attributed to Mark Twain but the sentiment almost certainly predated and absolutely certainly outlived him. Indeed, it is the core of what is taught in statistics courses. I sought the assistance of a well-respected quantitative researcher – one who discovers Truth statistically – to assist me with this small but mandatory element of my doctoral work. She started by saying ‘Tell me what you want to prove and we’ll figure out how to do that.’ And that’s a fact.) The argument my brother and I consistently have is whether aggregate Truth – the norm, the usual – is adequate Truth.
Data ‘holds knowledge that humans can’t know directly but must be extracted by a computer.’ The job of computers is to mine data for patterns which then inarguably define Truth, statistics on steroids. Which is then used predictively: if we do this, that will happen. Data is created by the recently arrived toy-boy data science, which Lapore says is associated with ‘late capitalism, authoritarianism, techno-utopianism’. This may be the basis for concerns the people who developed and delivered Artificial Intelligence have recently shared, that the Frankenstein that data science created – because it could and therefore should -- may indeed become the master. (This is not yet a fact. Certainly it is a mystery. Primarily its support comes from numbers, in particular economic prognostications, which, as we well know, describe reality in a way that blurs differences – e.g. safe and lethal fungi, or, say, economic inequality -- that are literally a matter of life and death.)
We’re in deep do-do if we buy that ‘only data tells because only data sells’. I’m with Dorothea, enthused about any – all?? -- kind of knowing that benefits the mainstream of people.
