Reification

What nature wants

“Nature gave us a sweet tooth because it wants us to eat fruit. So when you crave something sweet, that’s your body telling you what it needs.”

“Nature gave us” and “it wants” describe a long history of natural selection as if it were someone with a plan. The conclusion leans on that plan: if nature intended the craving, the craving must point at a real need. Restated without an agent (people whose ancestors liked sweet foods left more descendants, in a world with little sugar), nothing follows about what a craving means today. This is the clear-cut case, and it can lead straight into an Appeal to nature.

A company that values loyalty

“I’ve been here nine years. The company values loyalty, so it’ll look after me if there are layoffs.”

A company can have policies, and the people running it can value loyalty, but “the company” doesn’t remember anyone or feel grateful. The decision about layoffs will be made by particular managers, under particular pressures, possibly ones who arrived last year. Treating the organization as a person with a character makes a prediction about those managers look like a fact about a relationship. It’s less obvious than the first example because organizations really do act through their members.

The data says

At a planning meeting: “The data says we should cancel the Thursday evening class. Attendance is down 30% since spring.”

Attendance figures don’t say what to do. Getting from “attendance is down” to “cancel the class” takes choices nobody has mentioned: whether 30% is a lot for this class, whether the drop is seasonal, whether the remaining students have other options. “The data says” presents those human judgments as the verdict of an impartial agent, which makes them harder to question. This is the respectable-sounding version: the numbers may be accurate, and canceling may be the right call.

• Personified forces: nature, evolution, history, fate. “History will prove us right, so we don’t need to explain ourselves now.”
• Groups as persons: a company, a town, “the public”, given a single mind. Reasoning from a group to its members this way can also involve the Fallacy of division.
• Labels as causes: a word for a pattern treated as a thing that produces the pattern. “She’s impatient because she’s a Type A personality.” If “Type A” just means someone who acts impatient, it can’t also be the reason she does.
• Tools as authorities: “the data”, “the model”, “the algorithm”, “the science” treated as speakers with verdicts, when people chose the method and interpret the output.

Names. “Hypostatization” is another name for the same error, common in older textbooks. S. Morris Engel’s textbook defines the fallacy of hypostatization as “regarding an abstract word as if it were a concrete one” and notes that “Nature” is an especially favored subject. Alfred North Whitehead’s fallacy of misplaced concreteness (1925), “the accidental error of mistaking the abstract for the concrete”, is a close cousin, not the same thing: Whitehead aimed it at philosophy and science, at abstractions such as the idea of matter being simply located in space, taken as the concrete facts of nature. He wasn’t describing people who give abstractions intentions, and he was careful to say that the science built on those abstractions wasn’t “simply wrong”. The phrase is sometimes used loosely for reification in general.

Classification. Sources disagree on where reification belongs. Engel groups hypostatization with amphiboly among fallacies of ambiguity, since the error turns on how a word is taken. Robert Sinclair (2018) describes reification as a relative newcomer whose status as a fallacy isn’t well understood, and says the standard analysis treats it as a fallacy of presumption. If it is a matter of ambiguity, the ambiguity is in what kind of thing a noun refers to, not in grammar, so it sits closer to equivocation (one word in two senses) than to Amphiboly.

Talking about abstractions as if they act is normal and useful. It becomes a fallacy only when a conclusion depends on the abstraction literally having wants, judgments or powers.

• Metaphor everyone understands as metaphor. “The sea was cruel that night” or “the economy is sluggish” isn’t an argument, and nobody draws conclusions from the sea’s cruelty.
• Shorthand that can be spelled out. “The market punished the stock” stands for many investors selling. If the conclusion still follows once it’s spelled out, the shorthand did no harm.
• Groups that really decide through a procedure. A committee that votes, a court that rules, or a company that signs a contract does make decisions, in a well-defined sense. Reasoning from those decisions is fine; reasoning from the group’s supposed feelings or character is where it goes wrong.
• Models used as models. An economist’s “representative household” or a biologist’s “the gene for” is a deliberate simplification. It’s legitimate as long as nobody treats it as a creature with aims and draws conclusions from those aims.

Sinclair describes the standard advice as minimizing the causal agency assigned to abstractions, and discusses William James and John Dewey, who both stressed how indispensable abstract thought is.

The test: restate the claim in terms of the actual people, processes or measurements the word stands for. Does the conclusion still follow?

The market punished the stock

“After the recall, the market punished the company. The share price fell 20% in a week, so selling now locks in the loss.”

“The market punished” sounds like reification, but it’s shorthand that can be spelled out: many investors decided to sell, and the price fell. The conclusion about locking in a loss depends only on the price, not on the market having a motive, so nothing in the argument rests on the metaphor. It would become reification if the speaker went on: “the market is angry, and it’ll forgive them once they apologize.”

The committee decided

“The building committee decided to postpone the vote on the new roof until October, so we can’t order materials yet.”

Here a group is said to decide, and the conclusion depends on that decision. But the committee is a group that decides through a procedure: it met and voted, and its decision has a clear meaning that anyone could check in the minutes. Compare the company that “values loyalty”, where the conclusion rests on a feeling no procedure produced.

Grammar treats abstract nouns the same way as concrete ones. “Nature selected”, “the dog chased” and “the committee voted” all have the same structure, so the language itself doesn’t signal that one of these subjects isn’t the kind of thing that can act. Engel’s account puts the ambiguity in the word itself: a word that can properly refer only to an abstraction is taken as if it referred to something concrete.

Talking about abstractions as agents is also a real convenience. It compresses a great deal of information about many people, processes or measurements into one simple subject with a simple motive. That compression is useful right up to the point where someone draws a conclusion from the motive, which is easy to do without noticing that the motive was only ever a figure of speech.

• Spell it out. Ask who or what, concretely, is doing the wanting or deciding. “When you say the data says, which numbers, and who decided what they mean?”
• Check whether the conclusion survives. If it still follows once the agent language is removed, there’s no problem; if it doesn’t, the argument was resting on the metaphor.
• Separate the facts from the verdict. Numbers, histories and group decisions can be real even when the “verdict” attributed to them isn’t.
• Don’t reject the conclusion automatically. An argument that leaned on reification can still have a true conclusion. Maybe the class should be canceled; the data just didn’t decide that.