Arguing with our Ancestors

Madeline de Figueiredo used AI to recreate her late husband Eli’s voice and read his last email aloud. For a moment, the words seemed to have their speaker back. Later, the system offered advice so unlike him that she ended the experiment.

Her failed experiment leaves the next person something to examine before repeating it. It also raises a larger question: as technology gives the past new ways to speak, what do we say back?

We inherit traditions with the editing already done. We want the wisdom; we conceal the failed convictions, bad decisions, and apologies that helped produce it. More documentation does not automatically mean more candor. If my children inherit only my conclusions, they lose the process by which I learned to question them.

Unfortunately for my vanity, I have already published the evidence. I’ve always enjoyed writing, whether anyone appreciates it or not. I am, however, very open to being appreciated.

My blog has been public, private, public again, and nominally anonymous. Religion and politics supplied the controversy; family and employment supplied reasons to reconsider the audience. Twitter underwent similar adjustments. Eventually I abandoned it. I disliked who I was becoming in a place where having an audience could pass for having a point.

Some older posts still make me wince: a younger man, naïve, needlessly angry, and considerably more impressed with his conclusions than I am. I deleted some posts years ago. Now I would rather leave the evidence, add a dated correction, and explain which assumption failed—or whom I hurt. Keeping an old argument visible does not oblige me to defend it.

In a post dated 2008, I described the blog as something to pass down to my children, “to show them how you grew and changed over the years.” I still want that inheritance. A reader should be able to see what changed my mind—and whether my conduct caught up.

I wanted something similar during my chemistry PhD work at the University of Florida and later at the University of Kentucky. Published results were wonderful. But where was the repository for well-documented dead ends? What had people tried, under which conditions, and why had they abandoned it?

A failed reaction does not prove it can never work. It tells you what happened under particular conditions. Record the method and limitations, and another graduate student may avoid spending months rediscovering the same obstacle. Without that record, the first person learns something. The second person pays to learn it again.

Reading a business school professor’s case against business schools in the Times brought me back to the problem of curation. For cases developed with participating companies, Harvard’s own account describes managers reviewing drafts, flagging material they would prefer omitted, and authorizing publication. Other cases use public sources. But the mechanism deserves attention: the subject’s permission helps determine what students get to examine. History, subject to management approval.

Privacy matters. But when protecting a reputation means sanitizing a failure, the next person inherits the blind spot. We make decisions before knowing how they turn out. Edit away the uncertainty and you erase what could teach judgment—the process data the next person could use to engineer around those failure modes.

Jonah, for one, might have preferred his book to end one chapter early (Jonah 4:1–11).

The young are already arguing with the world we want them to inherit. In India, the Cockroach Janta Party grew from online satire into a movement over leaked exams and limited opportunities. By October, it was also challenging changes to voter lists. The cockroaches, inconveniently, were asking for more light.

In France, student protests this October drew tens of thousands, with clashes in some places. At a school in Saint-Denis, students described overheated classrooms, punishing schedules, and unequal educational opportunities. Looting is not going to help their cause, but dismissing them as radical is stupid.

The young can be a barometer of our conscience because they still take our promises seriously. They have had less time to build careers, identities, and respectable excuses around the gap between what we say and what we do.

I am old enough now to find the pressure gauge disrespectful.

Young people can also be passionately wrong. A slogan can masquerade as an explanation at any age. Their anger deserves attention; their proposed remedies still need an argument.

Marx and Bakunin give us a longer record of that struggle. We still fight over socialism, communism, anarchism, and capitalism, having inherited the vocabulary more reliably than the reading list. (We are lucky if the next generation even understands the labels.) 

Their doubts and false starts deserve some attention, too.

At nineteen, Marx wrote his father an unflattering review of his own poetry and philosophical projects. He had burned poems and plans for fiction. His delete button involved fire. The letter survived: a young man in love, exhausted, admitting mistakes and confessing he had read Jenny’s letter twelve times.

In 1842, Bakunin mocked men who had once loved freedom and now called their faded enthusiasm “experience.” He also thought there was little point talking to them because they would soon be dead. In 1873, he resigned from the Jura Federation and the International, citing age and health. His revolutionary convictions remained: the spirit was willing, but the flesh was weak.

Neither man supplies the reassuring journey from young radical to old moderate. Nor does moderation deserve an automatic certificate of wisdom. Sometimes experience corrects arrogance. Sometimes comfort learns to call itself experience. Keeping the record lets the older self correct the younger one—and lets the younger one cross-examine the old man’s story about his maturity.

I had hoped arguing with myself would improve my odds.

Social media can organize a movement, as the Cockroach Party shows. It can also turn argument into a revenue stream. Twitter trained me to produce a reaction; it gave me less practice showing how a thought changed. A screenshot can outlive its context. My early blogging was clout-thirsty, too. Yet an essay can keep dates, objections, and corrections together, giving a reader room to follow a mind through something, navel-gazing and all.

That is the inheritance I want: failed attempts beside later insights, confident assertions beside their corrections. A tool that puts two contradictory letters side by side might teach more than a synthetic ancestor dispensing polished advice. If it invents answers to fill the gaps, we lose the process again. The gaps belong in the record, too. 

Technology gives us new ways to preserve our lives and revisit those that preceded ours. I hope my ancestors continue arguing with me.

The Ten Commandments were written in stone, after all. Yet scripture also preserved Moses coming down the mountain, losing his temper, and smashing the tablets (Exodus 32:19). The story did not omit the golden calf to protect the congregation’s reputation.

Even a commandment carved in stone needed a story about the people who broke it.

---

Related Posts

• The many motives of blogging (August 1, 2008)
  An early reflection on blogging as memoir, the fear of forgetting, and whether writing might outlive its author. The ancestor question was already taking shape inside the debate over publicity, privacy, and deletion.
• Losing our humanity, one status update at a time (August 2, 2010)
  An earlier reckoning with digital permanence, the performance of an online self, and the right to outgrow our worst moments. Preserving a life’s record also requires room for forgiveness and change.
• Tell me what you think about your motivation (July 7, 2014)
  A candid account of leaving chemistry doctoral research, returning to teaching, and wrestling with motivation. The uncertainty stays visible, making the post a record of a life being reconsidered.
• Writers Tune the Zeitgeist (January 6, 2026)
  How enduring writing makes an age’s pressures and consequences understandable to other people. It extends the ancestor question into literature: what can a later reader learn by inhabiting an earlier struggle?
• Failure Modes of the Model (January 8, 2026)
  An examination of how a model can fail through rigidity, noise, or self-protection. It asks what makes a person capable of receiving correction when information alone is insufficient.