Clerestory

On Perfection

November 20, 2021

What does perfection mean to you?

Had you asked me this question a few months ago, I might have replied, tongue-in-cheek, that it’s a made-up concept that people use as an excuse not to start things. I’ve always been too slapdash to fall into perfectionism, as may be apparent from my patent lack of any perfections, in my work or in my life.

My own sardonic response reminded me of Ambrose Bierce’s The Devil’s Dictionary (1906). I wondered if he had an entry for perfection. In fact, he does!

PERFECTION, n. An imaginary state of quality distinguished from the actual by an element known as excellence; an attribute of the critic.

If I’m reading him correctly, perfection and excellence are imaginary, and critics imagine themselves to be perfect.

But what about true perfection? It’s not something I’d thought a lot about.

Spinoza, whom I wrote about yesterday, begins his section “Of Human Bondage, or of the Powers of the Affects” with this question. His explication is important. He illustrates how this word in particular, which was once concrete, became abstract over time, by steps. This relates to my recent writing on vision and abstraction, and also to a thread I’ve been working on about whether words were once more closely tied to the world.

Edwin Curley, a living translator of Spinoza, puts this footnote in the section: “As Hale White observed, it is important in understanding Spinoza’s analysis […] to realize that perfectus is simply the past participle of perficere, to complete or finish, itself a derivative of facere, to make or do.”

In other words, “perfect” originally just meant “complete.”

Here’s Spinoza:

If someone has decided to make something, and has finished it, then he will call his thing perfect—and so will anyone who rightly knows, or thinks he knows, the mind and purpose of the Author of the work. For example, if someone sees a work (which I suppose to be not yet completed), and knows that the purpose of the Author of that work is to build a house, he will say that it is imperfect. On the other hand, he will call it perfect as soon as he sees that the work has been carried through to the end which its Author has decided to give it. But if someone sees a work whose like he has never seen, and does not know the mind of its maker, he will, of course, not be able to know whether that work is perfect or imperfect. And this seems to have been the first meaning of these words.

Perfection is at first in the eye not of the Beholder, but of the Author (or, in this case, the builder). If I’m building a house, and I stop building, it is complete, therefore perfect. If you know what I’m up to (say, by talking to me), you’ll know when the house is done too. If I’m done, it’s “perfect.” If I’m still working on it, then it’s “imperfect,” which literally means “not complete.”1

Notice that he already qualifies this situation with “or thinks he knows.” Even in the concrete example that you can point to in the world, theory of mind has slipped in. I will call the characteristic of being able to point at something “indexicality,” after the index (the Latin word for fore-finger). This is important because it produces consensus rather directly. It’s the difference between “Look at that bird!” vs “Look at the state of the West.”

But after men began to form universal ideas, and devise models of houses, buildings, towers, etc., and to prefer some models of things to others, it came about that each one called perfect what he saw agreed with the universal idea he had formed of this kind of thing, and imperfect, what he saw agreed less with the model he had conceived, even though its maker thought he had entirely finished it.

Universal ideas form after the particulars. Possibly it requires lots of particulars. At the same time, more complex buildings arise; human constructions articulate and become more complex.

With universality and complexity come preferences. Could this be rephrased? I might write: “With knowledge and choice come good and evil” as abstract concepts.

Think too of Nietzsche’s idea that “good and bad” is not the same opposition as “good and evil.” I’ll come back to this, but to summarize, he thinks that the older distinction is between “good” (what people do) and “bad” (what they don’t do). The dichotomy between “good” and “evil,” as judged by some external authority or higher standard, comes, for Nietzsche, much later.

Nietzsche’s argument seems to me similar to Spinoza’s. “Good” and “perfect” just mean what exists, what has been done, what is finished. It reminds me of the refrain “And God saw that it was good,” which recurs in slightly different forms in Genesis 1, specifically 1:4, 1:10, 1:12, 1:18, 1:21, 1:25, and especially 1:31: “God saw everything that he had made, and indeed, it was very good.”

I wondered if this might make more sense as “completed” rather than a value judgement; after all, whose standard is God using? I found this page on the Hebrew word used in this sense, tov (normally translated “good”):

First, we recall a basic principle in Hebrew, in Semitic languages in general: Words always start out life as a concrete something, and only later take on a conceptual meaning, or become an idea. The original meaning of tov is not “good.” […] The meaning is more basic, something concrete.

The King James Version translates these verses, “And God saw that it was good.” […] But that’s not what tov means in the Creation story. The original, biblical, concrete meaning of tov is more along the lines of “well-formed,” “well-wrought,” “well-crafted,” and these meanings make sense in the context of Creation. “Good” is conceptual, value-laden and suggests values that play no part in the early biblical narrative. Tov as “good” is a later understanding of the word, perhaps derived from its original meaning but it is not the original meaning.

Interesting! This seems to line up very well with my intuition.

So together, universal ideas, complexity, and preferences lead to an abstraction of the notion of perfection. Now, suddenly, houses can be more or less perfect in comparison not just to one another, but to some higher standard. Furthermore, this judgement is not made by the Author, for whom something is “perfect” when finished, but a judgement of the Beholder, whose judgement is based on past observations.

Each person, through a process of inference, has arrived at an abstract notion (“universal idea”) of what completion means. Then “perfect” comes to mean not the degree to which something is “complete,” but the degree to which it conforms to an inferred abstraction. Sound familiar?

To me it seems like it must be related to Platonic forms. Plato is mistaking these abstractions for pre-existing objects, because they seem to form so automatically.2

Back to Spinoza:

Nor does there seem to be any other reason why men also commonly call perfect or imperfect natural things, which have not been made by human hand. For they are accustomed to form universal ideas of natural things as much as they do of artificial ones. They regard these universal ideas as models of things, and believe that nature (which they think does nothing except for the sake of some end) looks to them, and sets them before itself as models. So when they see something happen in nature which does not agree with the model they have conceived of this kind of thing, they believe that Nature itself has failed or sinned, and left the thing imperfect.

So far, “perfection” has involved two mistakes: First, we’ve changed the measure of perfection, replacing concrete completeness with an abstraction based on inference. Second, we’ve replaced the author’s estimation with the beholder’s.3

Here, Spinoza introduces a third mistake: We assume that nature operates like a human, which is a form of anthropomorphism.

Humans operate with ends in mind. They behave teleologically. I mentioned this tendency offhandedly in discussing the “Western” tendency to see the present as the telos of the past.

I suspect this is because reason itself comes from two sources which are much more teleological than nature is. These are:

  1. Immediate indexical reality, where if you drop something, it breaks, with simple causality that is readily learned. Animals must have this kind of reasoning.
  2. Social reality, where people act (or believe themselves to act) with goals and reasonably straightforward motivations. These are harder to learn but still possible; and possibly it is social pressure — i.e., the need to reason about the social consequences of actions — that allows anything like “abstract reason” to arise in the first place.

It is not that causality does not apply in nature. It’s just that simple causality is far less common than it is in the case of man-made things or situations — the glass that breaks when dropped, the person who retaliates when struck.

But the horse does not gallop — nor the human think — because of any goal. There is no simple causality as with an object in the hand, nor any intentionality as in social action.

Spinoza thinks this anthropomorphism is a mistake. Nature is not teleological. We only think it is because we lack the ability to apprehend things that there is no need (evolutionarily speaking) for us to apprehend.

This is at the heart of Darwin’s thinking, but you can see it already in Spinoza. Schopenhauer (who thinks nature is teleological) points also to Lucretius and to Francis Bacon as non-teleological thinkers. I hope to treat those other thinkers in the coming months.

I continue the discussion of abstraction here.


  1. It strikes me that this must be the sense still used in the imperfect tense; “I was thinking…” implies the thinking could still be incomplete.
  2. As Anil Seth argues in his new book, even perception/manipulation of objects in the world requires a process of inference. In this view, both “real objects” and “abstractions” are the phenomenological results of processes of inference. The only difference between the two is the degree to which the perception is “controlled” by the environment.
  3. Arguably, Plato makes these first two mistakes and then a third mistake that the abstractions are somehow more “real” than the instances that produced them.

Bryan Kam

I'm Bryan Kam. I'm thinking about complexity and selfhood. Please sign up to my newsletter or see more here.