Excel-sius
Our Father, Who Art in the Cloud
You will not be surprised to learn that your Director is not, by temperament or calling, a great enthusiast for the meetings. And yet it is one of those truths universally acknowledged that where two or more teachers gather together, a meeting will very often ensue. Austen did not necessarily have a staff briefing in mind, but she would, I suspect, have recognised the species.
On one such occasion I found myself in a room with the coffee going cold and a PowerPoint humming faintly in the background: that low electronic purr which suggests that something statistically ominous is about to happen. It did. At a certain point an esteemed colleague leaned forward and said, with the air of a trump card being placed gently but decisively upon the table: Well, the data shows…
Conversation paused. Heads nodded. The matter, it seemed, had been settled. I half expected incense. Once the data has spoken, there is very little left for mere humans to add.
Of course, what had actually come to pass was that a set of human beings had selected a series of variables, agreed upon a methodology, decided what to count and what to leave alone, and then produced a number. But that is rather less majestic. The data shows has a liturgical confidence that we designed a survey simply cannot rival.
We live, we are often told, in the age of data. It is collected, stored, mined, cleaned, visualised, monetised, secured, backed up, and occasionally leaked. We are encouraged to be data-driven, data-literate, and ideally data-compliant. Whole industries now exist to persuade unruly datasets to yield coherent narratives. The job title Data Manager would have baffled Aristotle, though he would, I suspect, have admired the ambition to systematise reality.
Perhaps the most striking emblem of our moment is the data centre. These vast, windowless buildings, low, rectangular, and faintly monastic, sit quietly at the edges of towns, humming with subdued intensity. They consume astonishing quantities of electricity and water to keep their servers cool. They exist for the sole purpose of storing, processing, and safeguarding information that, in theory, has no physical form at all.
The cloud, it turns out, is extremely heavy.
There is something faintly comitragic about the fact that our most abstract digital lives depend upon acres of land, miles of cabling, industrial cooling systems, and an energy appetite that would make earlier centuries blink. One wonders whether the metaphor of the cloud was devised in a moment of marketing genius: something soft, white, and heavenly, created to distract us from the distinctly earthbound warehouses quietly multiplying over there by the ring road.
None of this is to suggest that data is undesirable. On the contrary, it is indispensable. Human beings have always counted things. Ancient civilisations kept meticulous records of harvests and taxes. Medieval monasteries tracked donations and landholdings with an industry that would impress any modern auditor. Empires ran on census returns. Counting is not new. What does feel new, at least to your Director’s slightly creaking bones, is the tone in which counting now speaks.
Data has acquired a curious moral authority. It is no longer merely helpful; it is legitimising. A proposal supported by data feels serious. A judgment unsupported by numbers feels faintly irresponsible. We trust data in a way previous generations trusted tradition, testimony, or inherited wisdom. To say, I think, is to open the door of debate. To say, the data shows, is to close it.
Earlier societies were not free of error, prejudice, or dogmatism; far from it. But they tended to recognise more openly that knowledge arrived through human mediation. A farmer reading the sky was exercising their experience. A ship’s captain keeping a log was making observations shaped by judgement. A physician diagnosing illness relied not only on symptoms but on interpretive skill. Information was inseparable from interpretation.
The word itself is instructive. Data comes from the Latin datum: that which is given. It suggests something offered up to thought; a gift, not a decree. Data was once an invitation to consider, not a command to comply. Somewhere along the way, the invitation became an instruction.
Here might be a moment to pause for friend of the Column, Charlie Dickens. In Hard Times, Thomas Gradgrind insists that children be filled with Facts and nothing but Facts. Imagination and ambiguity are treated as dangerous indulgences. The world, he believes, can be rendered manageable if only it is sufficiently enumerated. Dickens’ satire feels uncomfortably contemporary. Gradgrind would recognise the data dashboard with a glow of vindication.
Yet Dickens knew, as all novelists know, that human beings are not reducible to units. You may count a classroom, but you cannot quantify a childhood. You may record attendance, but you cannot measure wonder. Gradgrindism fails because it mistakes the ledger for the life. It assumes that if you have the statistics of the school, you have the soul of the student.
T.S. Eliot once asked, Where is the wisdom we have lost in knowledge? Where is the knowledge we have lost in information? One suspects he would now need to add a further line about spreadsheets. For we would do well to remember that in the beginning was not the spreadsheet, but the Word: the Logos.
In the classical and biblical traditions, the Logos represents not just speech, but a cosmic order, an underlying reason, and a narrative meaning that precedes any attempt to count it. The Logos is the breath that gives life; the spreadsheet is merely the ledger that records the breaths. When we elevate the metric above the meaning, we risk forgetting that the story is the source and the data is merely the shadow.
The spreadsheet excels (your Director permits himself this modest pun) at answering the what. It is less comfortable with the why. However many data points we accumulate, meaning does not automatically emerge from aggregation. You can stack up an impressive quantity of whats without ever arriving at purpose. The ledger records; it does not interpret.
Plato, in the Phaedrus, worried that writing would create the illusion of wisdom without its reality: that people would possess information without understanding. It seems we have not entirely resolved the tension. Our current difficulty is not scarcity of information but abundance. We risk mistaking accumulation for insight.
This brings us to a particular kind of modern insanity. G.K. Chesterton, that master of the paradox, once observed that the madman is not the man who has lost his reason; the madman is the man who has lost everything except his reason. Chesterton understood that pure, cold logic, when divorced from the messy reality of human intuition, becomes a prison. An organisation that is purely data-driven is, in Chesterton’s sense, a mad organisation. It may be perfectly logical, it may have impeccable charts, and it may follow every trend line to its conclusion, but it has lost its mind because it has lost its context. It has become a giant, rationalising machine that can explain everything and understand nothing.
Your Director has sat in meetings (as near to the back as possible) where a projected graph was regarded with the solemnity once reserved for stained glass. A line trends upward and we all lean forward because the upwardness is self-evidently good and worthy of our rapt attention. Whether we understand what is being measured is almost beside the point. The numbers glow with an air of authority. Numbers are useful precisely because they simplify. They compress complexity into digestible form. They allow comparisons across time and place. In medicine, engineering, climate science, and countless other fields, careful measurement saves lives. The danger lies not in measurement but in misplacement. We may be asking data to carry a weight it cannot bear.
Our fascination with generative artificial intelligence illustrates the point perfectly. These systems ingest extraordinary quantities of human language and learn to predict the statistically probable next word. The achievement is remarkable; the outputs can be fluent, coherent, and occasionally even lyrical. But what they produce is pattern without experience.
Generative AI is Gradgrind on steroids. It is the ultimate Fact machine, capable of massive synthesis while remaining entirely oblivious to the truth, beauty, or suffering of the experiences it describes. An algorithm can summarise grief but it cannot decide when to say nothing. It can simulate the structure of conscience, but it has never lost sleep over a decision. It can inform you how often a sonnet deploys the word ‘love,’ but it has never misread a text message at midnight. You get the point…
Dear friend and frequent visitor to the Column, JL Borges in Del rigor en la ciencia wrote of a map so detailed that it matched the territory point for point and thereby became useless. There is a quiet warning there. Perfect representation is not the same as understanding. If we rely on the automated synthesis of data to tell us what to do, we are essentially asking a map to tell us what the forest feels like. Another great friend of the Column, Iain McGilchrist (Bingo!), also reminds us that representations are not reality. A map is indispensable, but no one imagines that the map exhausts the territory. You cannot smell pine needles or feel uneven ground by consulting a contour line. Data is a map: a powerful one, but it remains a representation.
This matters because we increasingly assume that if we accumulate enough information, wisdom will follow automatically. But wisdom involves judgement, context, imagination, and restraint. It is as much about what not to say as what to calculate.
None of this renders data suspect. It simply means that before a number appears, choices have already been made. Someone has decided what to measure, how to measure it, and what to ignore. To prioritise examination outcomes rather than intellectual curiosity, attendance figures rather than belonging, or productivity rather than purpose: these are not neutral acts. They are expressions of value, even when disguised as inevitabilities.
We should also probably confer a passing wave on Goodhart’s Law. This reminds us that when a measure becomes a target, it ceases to be a good measure. If we judge an institution solely by a single metric, behaviour will reorganise itself around that metric: sometimes helpfully, sometimes not. Hospitals can meet waiting-time targets without necessarily improving care; schools can raise certain indicators while narrowing the richness of the curriculum. The numbers improve, but the experience may not.
Kafka understood this very well. His bureaucratic worlds are meticulous in record-keeping yet opaque in purpose. Systems become so adept at optimising their metrics that they forget what the metrics were intended to serve. The file exists, the number is correct, and the procedure has been followed; yet justice is nowhere to be found.
And here the conversation acquires moral texture. Decisions about data shape opportunities, allocate resources, and influence outcomes. They determine who is counted and who is marginal. Hannah Arendt warned of the way administrative systems can dilute personal responsibility. When decisions are described as data-driven, accountability can become pleasantly diffuse. Your Director remembers a time when the devil made me do it served as a half-comic attempt to shift responsibility elsewhere. The joke depended on our recognition that the excuse was thin. We may have simply updated the language. The data made me do it sounds far more respectable; less sulphur, more spreadsheet, but it can function in precisely the same way: a polite displacement of judgement.
Yet to abandon data would be folly. Conspiracy flourishes where careful measurement is dismissed. The alternative to data is not wisdom but confusion. The question, then, is not whether to measure but how to remember what measurement cannot do. Perhaps the more fruitful inquiry is not What does the data show? but what does the data not show? What eludes capture? The craftsman’s intuition; the teacher’s sense that a room has shifted; the doctor’s quiet concern before the test results confirm it; the look exchanged across a classroom when something genuinely lands. These are the unmeasurables that give life its flavour and its meaning.
Of course, your Director is not immune to data’s charms. I must confess to a genuine affection for well-constructed data. There is pleasure in clarity and relief in pattern. A tidy table can feel like order wrested from chaos. But that emotional satisfaction is precisely why we must tread carefully. Numbers reassure us because they appear to reduce uncertainty. They feel solid.
But, whisper it quietly, they are only as solid as the assumptions beneath them.
As we continue into an increasingly quantified century, the servers will hum more loudly, the dashboards will multiply, and the cloud will remain impressively heavy. The task is not to dethrone the spreadsheet but to remember its place. It is an instrument, an extraordinarily powerful one. It can illuminate patterns and sharpen questions, but it cannot relieve us of judgement. It cannot tell us what ultimately matters.
Behind every dataset lies a series of human choices about what to notice. Behind every choice stands a person. And that person, inconveniently, gloriously, remains more complex than any column can contain. In the beginning was the Word, and in the end, it is our responsibility to ensure the story we tell is larger than the numbers we count.
Until next time, Happy reading / ordering your rows!
The Director’s Detritus #3
Did you know…? A single gram of DNA can theoretically store 215 million gigabytes of data. It seems nature perfected Big Data long before we built our first server, using a filing system far more elegant than any spreadsheet. We are walking libraries of a code we didn’t write; perhaps the most important information is already held in our bones. You never know…