We want to believe that there is nothing natural about numbers. These ciphers are our perfect projections, whom we marshal and archive, in endless right angles. Though they have pores and spines, the tables we fashion for them are not carcasses or skeletons, not birds shorn of feathers and skin. Believing that power to make tables is power to see deeply has done harm. And so, if I have acted to protect numbers, rather than subdue or cage them, I have done so not because I believe in their integrity or necessity. Instead I believe that human good depends on unflinching relationships with all that we meet and all that we summon, all that we eventually forget. The volatile markets of my career, awash in numbers trailing zeros by the hundreds, became their own reason for being. And all else these numbers authorized flooded the bodies of the willing.
By her own admission, Dara was never comfortable with statistics, the science of the quantitative sea. She took as easily as any scientist must to its basic terminology and could distinguish among obvious families (linear, logistic, Bayesian, quantile, ecologic, jack-knife) the way an historian can tell a Doric from an Ionic column. Her strength, as gleaned from a series of personality tests administered when she could barely find her way around the company compound, was catching discrepancies. (The company is an agglomeration of holding concerns that in the 70s succeeded in developing novel and profitable scientific products and there was much out of line at its inception.) With precision she could recall strings and nets of numbers, hold the relationships between types of numerical data in her head, and match these sets from location to location. In other words, she could say to herself “The document expresses relationships x and y on pages 25-60,” then repeat these phrases as she plowed her way through 400 pages of tables and charts.
Was this a form of parroting like or unlike analysis? It is difficult to know. Dara possessed a capacity for storing abstraction and decontextualized knowledge, carrying them from office to meeting like a tray piled high with china. She moved with grace, which for most of her career meant leaving fragile things undisturbed. Dara served the company in the way that made us uneasy: she scrutinized the work of dozens, becoming aware of their rate of output, the aspects of their jobs they relished and those they dreaded: the very forms their minds imposed on matter.
An early condition of the work, one she did not request but which we all understood necessary, was that she was never the one to confront the person or persons associated with the discrepancy. However, in the late years of her career, a vice president reversed the policy, and argued to his team that only Dara’s willingness to confront the perpetrator of error could affirm confidence in her finding. What lurked beneath this, an open secret, was that very few people above Dara could look at her work and be sure that she was correct. This was an unintended consequence of super-efficient operations. Often as not, there was only one person in a unit who did a specific job and rare it was that anyone around her understood what she did.
And regardless of whether you believed Dara’s skill to be superficial or deep, hers was a type of mind that the company president believed they had too few of. In fact, there was only her. Throughout her career, the personality tests seemed to have correctly predicted her fitness for a high rank and exacting, unfinishable work. You could try to dismiss her as a nitpicker, a wonk who teased her miserable existence from the hide of the group, fault-finding, sowing discord.
In truth, she was a reconciler. She made our documents whole, restored our parallel body to a flawless and nourishing state. She filled her assigned hours examining reports and documents, poring over strings of numbers and identifying the patterns these marks made. Like an oncologist or a garbage collector, she knew us in a way we would not know ourselves.
In her first month, she caught two quite large transcription errors— one by a laboratory module leader who was within weeks of quitting and one by an intern, who had been slated to return the following the fall; she also identified a spreadsheet entry mistake that had resulted in numbers and blank squares alternating down an implied line.
This glitch, Dara laughed to me, recounting the week she discovered it, looked like “a messy part on a girl’s pig-tailed head” so obvious she couldn’t understand how she was the first to see it.
“Not a ‘glitch’ strictly speaking,” I pointed out.
“Right. How can my whole job be based on precision of expression—in numbers or words—and I suggest a wrong is non-human in origin, when absolutely it is human?” She looked long at me and I smiled back.
I felt for her. I could see that if she didn’t find mistakes, she wasn’t doing her job. The early successes with the short-timing scientist and the lazy intern were crucial to her survival in the company.
Eventually, though, they hung like albatrosses around her neck. During the prosperous, placid periods of Dara’s thirty years with the company, it was sometimes hard to believe that only trivial mistakes were occurring or that people had learned to correct their own mistakes. When anyone expressed pleasure at the efficiency of the operation, there was a small feeling that Dara had missed something. The length of time between making an error and thinking about the existence of error is not fixed. We can realize we are wrong within minutes or, sometimes, it will strike us, years later, like a rotten tooth. People like Dara are always necessary, but that fact might be unclear until hope of repair is lost.
In one of those placid yet not remarkably profitable periods, there was talk of changing her job description. Think of a security guard. If the companies housed in the building are valuable, she too is esteemed. This company, our company, in the space of a year went from valuable to marginal and so the security guards—both those who stood at the doors of the building and those, like Dara, who stood at the doors of the data sets—experienced similar devaluations. People favored putting her in charge of proofreading and also having her train another vice president in her methods of observation and research.
But because she scored high in areas that most others did not, no one could carry her tray. A senior manager pointed out that Dara’s job was not to stop all mistakes—just the watershed mistakes, the errors that cascade through operations and create whole new low standards. Big mistakes come from small mistakes, though, a newer hire argued. If she is attentive on one scale, she could easily adjust her vision.
“There is no point,” Dara said, “in peering over people’s shoulders.”
“I’ve never done that,” she said to me later. I couldn’t tell then if she realized that I occasionally peered over hers.
She counter-proposed mining the company’s archives to examine abandoned data sets for information whose use was emerging or had not been recognized.
“If you feel I am not working enough, which I dispute,” she said, “let me observe more, not prosecute more.”
I was charmed to find Dara had gotten her way because it meant she would be in my office more often, mine being the data management services wing; I created the housing for the workers’ reports and assessments, a difficult job, merging financial services with the remains of several medical and bio-synthetic technology companies and their adjutant laboratories. There were documents only available in hard copy and gaps in the early digital material. I was one of the few left to explain those gaps.
When Dara entered the archive, she confided that though she knew she was looking for information pertinent to on-going projects, she couldn’t stop looking for mistakes. “And I find them,” she said, her teeth flashing just briefly from behind her bare lips. “Of course, I find them.”
I thought of the dogs trained to find bodies in the wreckage of cities after earthquakes. Dogs who grew excited around cemeteries and listless at schools and fancy fairs. Live bodies of no use to them.
A ledger that housed a too-uniform table, the values entered in the file all the same, all reading 5.75, concerned Dara. The error-ridden page was otherwise identical to one that housed normally varied data, numbers in all their seeming disarray. Until Dara unearthed the table, no one had thought about the date of the page where copying keystrokes seemed to have run amok. It was a Black Monday #23. (By some point there had been enough Black Mondays we began to organize them numerically.) The senior account officer, Lincoln, had seen the actual numbers arriving on #23 and realized that they would disrupt accounting throughout the organization. He had created a dummy page, a safe page, to which we could (and did) return at a later date for the purposes of reconciliation. Soon after though he had been diagnosed with skin cancer. He recovered, but it was during his extended absence that Dara had found the page and having no one around to tell her another story of its origin, she assumed it was a lapse. What she had seen was a scar, tracks of stitches that once held the body together, as it was laboring to heal itself.
When I explained this to her, she seemed relieved.
“It’s lovely,” she said. “I had not believed that anyone here would do such a thing.”
“Such a thing?”
“Offer proof that something had been wrong, attach their name to that wrong.” She stroked the page. “I had come here thinking it might be best to remove the page, but now I think it should stay, that we should make it known to others.”
“I don’t advise it,” I said. “For one, I don’t think Lincoln would like that.”
She didn’t ask for the second reason, which was good, because I only had one.
As this phase of Dara’s career unfolded, I took my last promotion. Sagging profits crept upward, new men and women re-shaping our company in a yet leaner and more financially desirable form. Dara eventually found real mistakes in the past as well as some ideas for the newly powerful partners to consider. Her relevance was tenuous, but not altogether vanished.
“How do you keep the numbers straight?” I asked, after she recounted one of those small coups to me.
“I don’t try to keep them straight,” she said. “I try to remember them. Like you recall lines of code, I would think.”
In seeing code, I saw an action, something changing. Sometimes, though, increasingly as the years went on and the company tipped and lurched, I saw nothing. It amazed me that Dara avoided such a fate.
A month into another economic downturn, she found a good-sized real-time discrepancy in what came to be known as the Forsgren case. In brief, a mid-career scientist who had transitioned into management fiddled with lab results to delay the launch of an adhesive until he found a way to invest in it indirectly. As with Lincoln’s Black Monday #23 safe table, Forsgren had created a series of dummy pages, only instead of repeating the same obviously wrong number, he had used a series of false numbers, numbers that at first seemed random. Dara, though, discovered an identical array of numbers in the archive, which proved to her that he had imported figures from an old project, incorrectly assuming that no one would be looking back or be able, as Dara was, to balance a great skim of numbers in her head. She came to me with the pages, seeming both sure that I would again explain the incident as innocent and that this time I would be mistaken. Her hands were still as they indicated matching numbers. I had no intention of explaining the coincidence in any way that excluded error as Dara was tasked to define it. In fact, I knew that what we had before us was a concatenation of errors: mental, moral, personal, professional. Yet, again, I advised her to stand down. There were very good reasons for keeping this mistake a secret. Forsgren, I pointed out, had ascended high enough in the company that no one above him would be sullied in such a matter.
“A reconciler like you should find it easy to understand the relationships between the gains and the costs here,” I said.
“Forsgren will cost this company its reputation, in the end, though. That’s big. That’s true.”
“You’re the reconciler,” I said.
By this point, Dara had to come straight to Forsgren, as if she were a trial lawyer and not a reconciler. Such confrontations were hard to watch. Forsgren’s face during the whole of the hour she presented evidence told us everything we needed to know. These copied numbers would not unseat him.
“Do you recognize these ten figures? 8086, 1709, 6880, 2000, 3879, 3887, 6117, 52009, 56783, 1000087.”
Dara recited the figures from memory, expressing the importance of position and magnitude with the rise and fall of her voice. To her the numbers were features of faces; quantification never seemed so fragile, like a beak ripped unceremoniously from a darting head and pressed bloody onto a snake or dog whose vitals had been similarly amputated. Her voice, those vibrations emanating from her graying head, gave the numbers a force, a life, a particular thing like self.
Yet Forsgren did not hear this. To him, the numbers were as dead as the day he found them and drafted them into the service of a strange master.
For a minute, I thought that Dara had won. Glimmers of comprehension flashed here and there in the boardroom. Perhaps had she been given one more chance to read the table aloud, the outcome might have been different.
But then Forsgren himself spoke up, his voice even and rich and neat, as he folded the first page of Dara’s case file into quarters. He suggested that the principals all receive their own copy of the files in question and he, for his part, would make his own notes available to the firm.
It is not often that I dislike being right. Watching Dara fold her arms in anger and then in submission, I realized that there was only one human problem left that I cared to solve.
By the end of Dara’s trials, she three times withstood pressure to resign. That tray of fragile things she carried with such poise for more than twenty years, that knack she had for holding information in precise constellation, those things she kept safe, kept her, if not safe, then at least on the payroll. There were a few old hands like myself who still respected Dara as one respects the last of any species. The times she had re-sequenced a table, ordered numbers as they needed to be ordered, she had been the company’s lungs.
But to the new guard, she was a nuisance, a worker unable to pull her weight. At first the vice presidents accepted her recommendation that all of the projects Forsgren oversaw should be revisited. By the end of the Forsgren case, when Dara had been reversed and, according to Lincoln himself, shown to be in error, individuals and teams shut their doors on her with impunity. In order to fill her now dwindling time before retirement, Dara retreated to the archive. I watched her there because I could not help myself. I worried for her. She hadn’t seemed this lost since the first era of company serenity, when it seemed she did her job too well.
That is why I was the only one to notice her taking files from the building. Building security personnel were instructed to watch out for theft, of course. As people’s hours stretched and they brought their equipment back and forth and as physical files became rarer, security’s job, like Dara’s, became more difficult. What was normal in this night-shifted, altered world? Dara, now in her late 50s, was five years from retirement. Could she resurrect her own reputation, as she seemed to be trying to do?
To think of oneself as useful is hubris. To make oneself useful is gospel. Dara was still trying, despite everything, to be useful.
“Do you know,” I asked her one night as she was tucking manila folders into her leather satchel, “if they still give the same personality tests they did when you were first taken on?”
“It’s funny you should mention them,” she said. “There’s a giant stack of the old paper tests back there that somebody forgot.”
I had no doubt she had examined them; in the course of things, she must have. So she must have encountered herself as a younger woman, the mind and hand that answered questions in that unusual way.
“I have often thought it would be an interesting exercise to write a personality test,” I said.
“Tests that seem to be only about memory, about keeping figures arranged, not losing them in a sneeze or blink. Like those carnival games, the ones where you maneuver a mechanical claw into a pile of toys and if you manage to hang onto one, it’s yours to keep.”
I nodded, fuller of sympathy than I needed to be, as she put another file into the already strained satchel.
I said. “There’s the test then. Just have all the new hires try their hand at the ‘Arm of Truth.’”
After that night, we talked more frankly. About retirement, mostly. About horse farms and jam-faced grandchildren, luxury cruises where you keep circling the same deck looking for a whale in the distance. I also wondered aloud about the scientists and accountants and statisticians and quants who had written the tables and reports that slept around us. Dara and I knew that these men and women had experienced longueurs when puzzlement gave way to giggles. That they listened some days to the sound of each other’s breath, rehearsed Nobel Prize acceptance speeches in their heads, rehashed arguments with former professors, never quite winning, though in fantasy, the fantasist should win.
“It’s impossible now to take responsibility for a piece of knowledge, to take one column of numbers and know it,” I remember her saying. “The only way to succeed in a company like this is to act like you command the whole structure.”
“You know what else?” she asked. “I now know it’s impossible to convey to anyone what I know about the company. The time for making a discovery can pass. An aspect of Newtonian physics, Galenian biology, Aristotelian genetics, Mendelian botany. Even a minor dialect of Goth matters. But they are all pointless after the time for discovery is past.”
Dara had made the most of an exile, for all of a career she had tried, and for a time she was glad to see the world more fully, to overcome the sense that there was nothing in these ruins. But when confronted with naming what she found, she faltered. If there is no name and there is no place, then there is no thing. Her knack for carrying knowledge intact was of a higher order than I supposed. She was constantly reconfigured by these delicate projections; and the deep changes in her also marked me.
At first it felt wrong to pick back through files Dara pulled. But the day after she retired, I began.
Until I started, I did not realize that a part of my mind had been recording Dara’s ways all along, unaware of itself. Yet it took me a long time to find all of her traces, almost two years, for mine is a different sort of mind from hers. To see the patterns she had once seen and now imposed, I had to learn and re-learn how to make the numbers breathe as I heard her do at the Forsgren meeting. But the thrill of learning, of seeing a complexity in the world normally closed to me, kept me alert and able.
One by one, I found Forsgren’s figures where Dara had grafted them onto tables. Seeing the “08” and “09” sequences, I surged with affection, for her and them. Her choices were not random. In vandalizing this archive, she had sketched the shape of a better company, one in which accounting programs were stalled to give lagging competitors time to make enough profit and remain viable through the end of a bad fiscal year; one in which if people needed time to understand the full consequences of their decisions, they got it; one in which there was no price for saying, “I don’t know,” or asking “Why?” Her reformed company scrutinized itself, shored up its parent structures, insured the persistence of novel forms because their value was as yet unknown. Her work was crude in places; she used exclamation points where a series of subtle modifiers might do. But she did well with her unheralded imagination; until the end of her career, she grew.
Abandoned data sets are like ruins of ancient civilizations. We can’t quite know which lane buzzed with children or the alley that filled first when the drum corps struck up and a martial rhythm began snaking its way along walls and rooftops. Our insistence that it matters, for a while at least, is enough. The immensity of the reconciliation task and the minds of the reconcilers mirror each other over time. It might be enough to think of this as a question about choice and time. A mistake made near the beginning of life can’t help but portend. A mistake made near the end of life cannot predict much future, but must instead be regarded as an anomaly or force a reinterpretation of most of what led to it. The latter, so obviously arduous, is often avoided, even if it is the only course that gets at the truth. Sometimes we make distinctions among truths, as we esteem or scorn civilizations.
I retired two years after Dara, which according to Lincoln and the others, was long overdue. I had provided a house for numbers; Dara ministered them, remembered them when no one else would, in their ugliness and rigidity, in their peculiarity and mysteriously increasing weight, and in the end, she upheld them, as a form of ideal beauty.