Wordy Weapons of Is-Ought Alloy

[Note: I have a nagging feeling I’ve spent a few thousand words spelling out something completely obvious. Still, I hope there’s value in actually spelling it out.]

It depends on what the meaning of the word “is” is.

—William Jefferson Clinton

He was right, you know.

A few months ago I argued in The Big List of Existing Things that the word “exist” has no consistent definition in real terms. Therefore, whether a thing exists or not isn’t a well-defined question without a specific context telling us how to interpret the claim.

At the end I said that while “exist” doesn’t have a clear meaning when relating to reality itself, it does when relating to our models of it, our maps:

On an abstract level — call it map-speak, or pseudocode in programming parlance — asserting that these things do or do not exist has a clear single meaning: “this thing should be represented by a token in our map of reality”.


Even a single statement can be compiled in different ways. Note that the meaning “this thing should be represented in our map” can be rephrased into “a map representing this thing fits reality better than a map not representing this thing”. Its a claim of fit between reality and a map.

The fit is bidirectional. It has two degrees of freedom. You’re making two different claims at once, in a kind of superposition. We can see this by holding one degree of freedom fixed at a time:

If we hold the map fixed, we make a claim about reality. In the Kurdistan example, we can take the local definition of existence as a given (“is a recognized state with functional sovereignity”) and say that empirically, Kurdistan is not that. It’s not in the list of such things.

If we on the other hand hold reality fixed, we make a claim about maps. We take concrete facts about Kurdistan as a given and say that it exists. By doing that we insist that “exist” in this context (to be a country) should mean something that includes Kurdistan, like “being a geographical area dominated by an ethnic group with common culture and identity that we think deserves self-determination”.

Normally, we don’t specify what end we’re holding fixed and in many cases we don’t even know (hence “superposition”). If you say “Kurdistan doesn’t exist” in normal conversation you are (until you specify further) both saying something about the properties of Kurdistan and what you think it should mean for a country to be considered existing (in this case, some way that doesn’t include Kurdistan).

“Exist” is only a special case. In general, words don’t map onto reality perfectly (the world itself isn’t implemented in the kind of “code” we use to describe it), so it’s not possible to say anything about reality without making a choice about how to represent it[1].

If I say “that guy is fat” I don’t just describe the physical form of a person. I also make a statement about how “fat” should be defined: in a way that includes him. And if that wasn’t enough, I also implicitly claim that “fat” is a valid concept that can and should be used.

This bridges the old division between “is” and “ought”. Not in the sense that you can derive an ought from an is, but that every claim is an alloy of the two.

That’s not a bad thing in itself, the problem is that we can use a token to represent abstract, complex and value-laden concepts as easily as concrete, simple and neutral ones. That means that just looking at the form of a statement and representing it in our minds tells us almost nothing about how it relates to reality. Subjective and objective complexity can be wildly different, and we can’t intuitively tell if a claim is complex or simple, clear or ambiguous, a matter of fact or feeling, is or ought.

We rarely find “isses” and “oughts” in pure form. Alloys are their natural state, and the proportions vary: claims range from simply factual to pure expression of judgment. Factual claims result when the meaning of a word is so well established that there is essentially no serious disagreement about its proper use in a given context. This applies to most technical and scientific terms and many social facts that are actually defined in terms of human concepts, like the names of currencies or the winners of sports tournaments[2].

In those cases, applying a label is so close to a pure statement of fact that the “ought” element is negligible. It’s an alloy only trace elements away from being a pure substance[3]. Near universal consensus[4] makes it possible to say that some words have “correct” definitions and can be used to state facts.

For illustration, here are some nice facts:

Dolphins are mammals
Spain is a monarchy
Ganymede is Jupiter’s largest moon
The Toronto Maple Leafs is an ice hockey team
Proteins are chains of amino acids
A hexagon is a six-sided shape

Unfortunately there’s a cognitive flaw that makes us act as if the definitions of words are written into the fabric of reality, so that we can just take our idea of a word for granted. It’s as if our minds have disconnected the mechanisms that produces correct definitions (consensus)  from our subjective sense of correctness[5]. This makes us underestimate how common it is for disagreements to arise from different ideas about the validity and proper use of words.

Tools and weapons

Words are tools. They can communicate information. They are also art supplies, because they can express and induce feelings and impressions. And they are weapons. They cast judgment, invite sentiments, and shape the way we see the world: what seems real and what doesn’t, what we notice and what we don’t, what we celebrate and what we condemn.

The tools and weapons available in a society determines quite a lot about it, and that extends to verbal tools and weapons quite well. So we fight over them. We want there to be tools that make it easy to do what we want done, and we dont’t want there to be tools that make it easy to do what we don’t want done.

A reference to 1984 seems like the obvious thing to put right here. Oh well, if you insist:

Don’t you see that the whole aim of Newspeak is to narrow the range of thought? In the end we shall make thought-crime literally impossible, because there will be no words in which to express it. Every concept that can ever be needed will be expressed by exactly one word, with its meaning rigidly defined and all its subsidiary meanings rubbed out and forgotten… The process will still be continuing long after you and I are dead. Every year fewer and fewer words, and the range of consciousness always a little smaller. Even now, of course, there’s no reason or excuse for commiting thought-crime. It’s merely a question of self-discipline, reality-control. But in the end there won’t be any need even for that… Has it ever occcured to you, Winston, that by the year 2050, at the very latest, not a single human being will be alive who could understand such a conversation as we are having now?

You need not go that far. In a Black Mirror-manner long before Black Mirror existed, 1984 scares us by taking a real issue and turning up the volume to “deafening”. Newspeak is not entirely different from real languages, and it matters a great deal also in them what words are adopted and what they’re set up to mean. This is especially true for powerful, strategically important Stalingrad-grade words like “free speech“, “democracy”, “rape“, “disease“, or the tactical nuke of verbal armaments: “racism”.

These words are powerful because they come with built-in value judgments, and their use forms and strengthens associations in our collective consciousness.

To agree that a word accurately describes a thing is to agree that a whole set of attitudes and associated actions should be applied to it. What is and what isn’t “theft”, “hate speech”, “torture”, “terrorism” or “eugenics” is one of substance, sure, but perhaps more importantly one of consequences. If accused of a crime there may or may not be something bad you have in fact done, but in the end it is the label that sticks — guilty or not guilty — that determines what happens to you. Other labels come with less stark consequences, but with consequences they come.

The lyrics to Wonderful performed by the “wonderful” Wizard of Oz in the musical Wicked is a wonderfully (sic) cute illustration of the power of labels:

A man’s called a traitor or liberator.
A rich man’s a thief, or philanthropist.
Is one a crusader,
or ruthless invader?
It’s all in which label
is able,
to persist.
There are precious few at ease
with moral ambiguities,
so we act as though they don’t exist.
They call me wonderful, so I am wonderful.
In fact it’s so much who I am, it’s part of my name,
and with my help, you can be the same.

(Couldn’t resist that, I’m a sucker for clever rhyming.)

Labels sort individual cases into predefined buckets for easy parsing and propagation. This is how minds work. This is efficient. This is a good thing. A necessary thing. We save effort when we think and we save bandwidth when we communicate by using preset concepts instead of preserving full complexity and nuance. We have not the time, not the inclination, and not the processing power to communicate using anything more than a caricature of reality.

Given all this, of course we’d like to control what labels are in use and what connotations and associated actions they have. We’d be stupid not to try.

Arguing semantics are often derided, but it shouldn’t be. Telling someone not to argue semantics is to imply that they should accept whatever vocabulary is given to them, i.e. accept to have the terms of debate dictated to them. That’s often tantamount to begging the question, since a lot of public discourse is dedicated to shaping the meaning of labels and getting them to stick to certain things (and to resisting and disrupting your enemies’ attempts to do this). That’s PR. That’s rhetoric. But not traditonal, “focused” rhetoric where a man in a toga tries to convince the senate to launch an invasion. This “unfocused rhetoric” isn’t about some specific decision, it’s about influencing the background memetic environment to be more favourable to your interests.

In other words it’s not pure “ought” as in “we should do this”, because it gets elements of “is” when it tells us to adopt a particular model of reality and its associated vocabulary.

Lead by example

Words get their meaning from their use. That is, we infer meanings from hearing others talk. So the natural strategy is to use words the way we want them to be used and hope that others go along.

Sometimes this is too obvious for anyone to miss. Not often though, since we’re so bad at grasping the disconnect between words and reality intuitively. It takes a lot for it to reach conscious awareness: claims need to be extreme and/or highly familiar.

A few examples:

Abortion is murder
Taxation is slavery
Property is theft
The personal is political
You can’t be racist against white people
Meat is murder
Oral sex is not sex
Marriage is between a man and a woman

Note that these contain virtually no actual information about the thing they’re supposedly about. Instead they use pretend factuality to tie concepts and labels together (or keep them apart). They’re the opposite of facts. They’re slogans.

Obvious slogans like these aren’t the most destructive type of claim. Stuff breaks, sure, but it’s a clean break. More ambiguous fact-slogan alloys have nastier failure modes, and they are the rule more than the exception. Think about claims

that women are oppressed
that Russia is not a democracy
that socialists don’t understand economics
that the new Star Trek series is not Star Trek
that we have free will
that the unemployed are lazy
that IQ tests are racist
that healthcare is a right
that homosexuality is a choice
that alcoholism is a disease
that fruit is healthy
that it’s selfish to want to have children
that it’s selfish to not want to have children
that progress is stagnating

that they support equality of opportunity
that consciousness is an illusion
that automation threatens jobs
that freedom of speech does not equate to freedom from consequences

that gender has nothing to do with sex
that someone is a climate change denier
that someone is an asshole/creep/bitch/cuck/misogynist/feminazi/whatthehellever
that something is rigged/organized/a conspiracy
that something isn’t proved
that something is ruined

that something is socially constructed
that something is or isn’t normal
that something is determinist

that something is unfair
that something is a mistake
that someone acts entitled
that… well anything, I could keep writing this list until the end of time, live out the rest of my days trapped in a Borges short story

With these, the most dysfunctional brand of shitstorm results because of how difficult it becomes to understand what the hell we’re even talking about. There is no consensus on the meaning of these words (or phrases, the mechanics are the same), but that isn’t obvious enough to be common understanding (“common understanding” in the sense of “common knowledge”).

These short, punchy strings of words refer to complex, ambiguous and poorly defined families of claims, but their linguistic form tricks us into thinking they’re simple and have simple answers.

So we interpret our preferred way, think that is the only way, and then go on to pretend to fight over facts[6]. I wrote about this the first time in Case study: The War on Christmas, in a passage that this whole post is more or less an elaboration of:

An encounter with an ambiguous yet controversial-sounding claim starts with an instinctive emotional reaction. We infer the intentions or agenda behind the claim, interpret it in the way most compatible with our own attitude, and then immediately forget the second step ever happened and confuse the intended meaning with our own interpretation. This is a complicated way of saying that if you feel a statement is part of a rival political narrative you’ll unconsciously interpret it to mean something false or unreasonable, and then think you disagree with people politically because they say false and unreasonable things.

We kind of have to do this, don’t we? Otherwise we’d have to admit to ourselves and to each other that because people disagree on the proper use of words (and on the validity of narratives when applying the same model on a larger scale), there is no consensus, no objective standard and thus we can’t justify ourselves because  “I’m trying to gain the upper hand here!” isn’t a compelling argument.

Wish: It should be standard practice to say “I reject your vocabulary”, instead of “you’re wrong”, “that’s just your opinion”, or whatever. It would force the conversation into one of two paths: either the point is restated in more neutral terms, or it becomes clear that a substantive point is not, well, the point, and that the goal is to impose vocabulary. Then the conversation can either continue in a constructive manner or end promptly.

As I said: wish.

Wrap-up: The moral philosophy of factual-looking claims

Separating is from ought and reaching for the real facts is a preoccupation of philosopher-scientist-type nerds, not most others. Ordinary language reflects this. Regular people who argue about what seems to be the facts are arguing about various practical issues, all of which are differently proportioned alloys of:

what the world is actually like


what certain words should mean.

The alloys are questions of the kind we actually debate, which is:

How should we model things?

This has different interpretations in practice, one is:

What ideas should influence how things are done?

Which can be extrapolated into:

How should things be done?

Which is what gets people riled up. Questions about how things truly are that don’t have social implications don’t turn into hot-button issues, and only specialist nerds care about them. When Sheldon and Leslie on The Big Bang Theory have heated arguments about the validity of string theory it’s supposed to show how weird they are.

Weirdo nerds differ from most others by not using that interpretation. Instead,

How should we model the world?

is interpreted as

What system of concepts capture the structure of reality with the most technical accuracy?

which is an extrapolation of

What are the concrete facts?

Which one of these interpretations you prefer[7] says quite a bit about you. It says whether you favor model consequentialism that evaluates models based on the likely consequences of their adoption, or model deontology that evaluates them based on their technical accuracy. The two focus on different causal properties of models when arguing for or against their adoption: they reflect reality and they shape reality. Which is more important?

Let’s just say that this is something people disagree about, and I expect it to relate to other psychological and ideological differences, like being concerned about reliability and predictability of rules and norms vs. a more the-ends-justifies-the-means attitude.

Aren’t we missing a standard ethical model? Yes we are.

How should we model the world?

also means

What ideas should I adopt?

which means

Who am I? (and who am I with?)

The last is model virtue ethics, a mindset that evaluates ideas based on whether believing/endorsing/supporting them is something a good person does, according to the standards of one’s community.

Some suggest that this is how our belief adoption mechanism actually works most of the time.

The traditional taxonomy from moral philosophy where consequentialists go up against deontologists and virtue ethicists can be applied quite easily to supposedly fact-type beliefs as well. And like in actual moral philosophy, they should be considered complements, not substitutes. Partial narratives, if you will.

All this is not to say that is and ought are not in fact separate or that Hume’s problem is solved. It’s not, of course. It’s just that if you treat them as separate things when trying to understand how humans think and argue, you’re gonna have a bad time. It can’t be treated as a basic axiom of thought. It’s advanced, technical stuff.

• • •


This becomes less of an issue as we move into hard science vocabulary, for complicated reasons having to do with compression and the use of “reality-matching concepts” (which I’ve been meaning to finish a post about any day now).

Note that those are the result of decision procedures whose function is partly to remove ambiguity.

Not zero, negligible. I’m not peddling any naive absolutism here. Non-naive, provisional, local as-if absolutism is my game.

I’m glossing over something here, namely that there are different reasons why there is consensus (and that some of them give rise to more “correct” definitions in a sense, but that’s outside the scope of this post). Formal decree is one, power-relations discourse is another, and painstaking study of the real world is a third. Let’s just say that different philosophical schools grab different parts of the elephant here and as a result have different ideas about the nature of truth.

This is possibly a result of us living in weird, unnatural conditions where not pretty much everyone we interact with is on the exact same page wrt to the meaning of cultural signs.

On an implicit, half-conscious level we do get it, but it doesn’t quite reach full awareness, and it certainly does not become common knowledge. This is likely an example of what Kevin Simler and Robin Hanson discuss in Elephant in the Brain: a self-serving behavior we remain unaware of because it helps us deceive others to also deceive ourselves.

Like with other such dichotomies, we don’t use just one or the other. We prefer one and use the other as an afterthought to adjust when needed.

Did you enjoy this article? Consider supporting Everything Studies on Patreon.

3 thoughts on “Wordy Weapons of Is-Ought Alloy

Leave a comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s