Slippery knowledge

 THOUGHT OF THE DAY:     
"No power and no treasure can outweigh the extension of our knowledge." - Democritus, from John Owen's 1881 Evenings with the Skeptics, p149. 
Sounds good, but not sure that it says anything, and if so, not sure what it says is true, since learning can be outweighed and, besides, can be forgotten.

Yes. It sounds profound, but once you poke it a little, it becomes rather slippery.

I think Democritus is making a value judgment, not really stating a proposition that can be tested. “No power and no treasure can outweigh the extension of our knowledge” means, roughly, knowledge is a good whose value exceeds the value of worldly possessions or political power. But that leaves the crucial question unanswered: why?

And your objections expose two weaknesses.

First, knowledge plainly can be outweighed by other things. If I could acquire a great scientific discovery only by causing the death of a person, I don't think most people would say, “Well, knowledge wins.” Even less dramatically, there are circumstances where comfort, friendship, freedom, security, or simply enjoying an afternoon might reasonably be valued more than learning one additional fact.

Second, knowledge isn't necessarily permanent property. If I learn something and then forget it, the extension of knowledge has disappeared for me. Democritus might reply that the knowledge has entered the general human stock and therefore remains valuable, but then the quotation is really about civilization's cumulative knowledge, not the individual's knowledge.

There is, however, something interesting underneath it. The quotation becomes much more defensible if “extension of our knowledge” means the expansion of humanity's understanding of reality. Power and treasure are means of controlling or possessing things; knowledge is a means of understanding what is actually the case. And knowledge can be passed on, corrected, accumulated, and used by people who haven't yet been born.

But even then, “no power and no treasure can outweigh it” is rhetoric. It isn't an argument.

In fact, I suspect the quotation's appeal comes partly from its being unfalsifiable moral poetry. Almost everyone who values learning can nod at it, because it doesn't specify what knowledge, whose knowledge, or what it should be weighed against.

So I'd give it a rating of excellent sentiment, dubious philosophy.

And your point about forgetting is particularly good. Perhaps the more defensible version would be: “The capacity of humanity to extend and preserve its knowledge is a treasure greater than any particular treasure.” That's a considerably more modest claim—and harder to knock down.

We don't know what human culture has known in the past and lost.

Exactly. And that makes the claim about knowledge being an accumulating treasure considerably less secure.

Human knowledge isn't a steadily rising pile. It's more like a palimpsest: some things are added, some are corrected, some are forgotten, and occasionally something that was lost is rediscovered.

We know of plenty of lost knowledge simply because later people left evidence that it once existed. But there is no way to make an inventory of what has disappeared without a trace. A technique known to some ancient craftsman, a medicinal observation, an astronomical record, a mathematical method, an oral tradition—all could vanish when the people who knew them died.

And there is an especially interesting asymmetry here: we know what we have rediscovered, but cannot know what we have never rediscovered. The unknown unknowns are precisely the lost knowledge that leaves no record.

So even the grand Enlightenment-ish idea that humanity is continually accumulating knowledge needs qualification. We may be accumulating net knowledge over very long periods, but we are also continually losing it.

That actually makes Democritus's statement more interesting to me. Perhaps the real treasure isn't simply knowledge, but the institutional ability to preserve, transmit, criticize, and rediscover knowledge. Libraries, writing, science, education, archives and traditions are valuable partly because they reduce the rate at which the human race has to reinvent the wheel.

And even those aren't foolproof. The Library of Alexandria is the cliché, but the broader point doesn't depend on that particular story. Every generation inherits a selection from the past, not the past itself.

So I'd be even more suspicious of the aphorism now. Democritus says, in effect, knowledge is the greatest treasure. But history gives us a rather humbling rejoinder:

We don't even know how much of the treasure we've already lost.

How many languages are extinct, and with them ways of thinking and literature?

A great many—and the number depends heavily on what counts as “extinct.” There is no reliable count of all languages that have ever existed, because most left no written record.

For languages that are documented, though, the loss is enormous. Ethnologue currently lists roughly 7,000 living languages, while linguistic databases record thousands that are extinct or no longer spoken as native languages. And many more are endangered, sometimes with only a handful of elderly speakers remaining.

The really important point is the one you're making: a language isn't merely a code for translating thoughts from one language into another. It embodies distinctions, metaphors, classifications, jokes, stories, poetry, names for plants and places, ways of describing relationships, and habitual ways of organizing experience. When the last speakers die, some of that can become literally unrecoverable.

Literature makes the loss even more obvious. A written language can leave us texts after the language has died—Sumerian, Akkadian, Classical Egyptian, Gothic, Old Norse, and many others. But if a language was primarily oral, its literature can disappear almost completely when its speakers do.

And there is a particularly poignant distinction: we can preserve a recording of a language without preserving the culture that made it meaningful. We might have the vocabulary and grammar, yet never recover all the associations, allusions and cultural knowledge that a native speaker would have possessed.

So this gives Democritus's maxim a rather nasty complication. Human knowledge may be expanding overall, but the set of things humanity knows is not monotonically increasing. We are simultaneously discovering things and losing things.

In fact, every extinct language represents a tiny experiment in human cognition and culture that can never be repeated. We don't merely lose words. We sometimes lose a way in which human beings once made sense of the world.

When Mozart (a stand in here for all creators) died, all but the most superficial knowledge of how his works came to be died with him.

Yes—and that is an even stronger challenge to the idea of knowledge as something we simply accumulate.

We possess Mozart's scores, letters, anecdotes, and some accounts from people who heard him. But we don't possess Mozart's knowledge of Mozart. We don't know what he heard internally when he composed, which alternatives he considered and rejected, what he could improvise and then never write down, how much was conscious calculation versus instinct, or precisely how his musical imagination worked.

And that distinction matters. A score preserves the result, but not the generative knowledge that produced it.

It is rather like having the source code of a program while losing the programmer's entire mental model of why it was written that way. We can analyze the code and infer quite a lot, but some of the original knowledge has simply vanished.

This applies to virtually every creator. Shakespeare's plays survive, but Shakespeare's knowledge of how he constructed a particular scene doesn't. Newton left an extraordinary amount behind, but Newton's working knowledge—the things he knew but never considered worth writing down—died with him. A great craftsman may leave an exquisite object while taking the technique for making it to the grave.

And there's another wrinkle: even Mozart himself probably couldn't have completely articulated his knowledge. Some of it was tacit knowledge acquired through years of practice. He could do things he could not necessarily explain. That's one reason apprenticeship and imitation are so important in the transmission of skills.

So perhaps the most significant thing lost at a person's death isn't the information they have recorded. It's the enormous quantity of unrecorded, tacit, embodied knowledge that existed only because that particular person existed.

Which makes your original Democritus quotation rather ironic. Knowledge may be extraordinarily valuable, but it is also extraordinarily perishable. The human race doesn't inherit the whole intellectual contents of its dead; it inherits the fragments they managed to leave behind.