The Nominalization Fix: Turning ‘Make a Decision’ Back Into ‘Decide’

Read this sentence out loud: “The team will make a determination regarding the implementation of the new policy.” Now read this one: “The team will decide how to roll out the new policy.” Same meaning, but the first one trudges and the second one walks. The difference is a quiet habit called nominalization, and once you can see it, you can’t unsee it.

Key takeaways

• Nominalization turns a verb or adjective into a noun, then props it up with a weak verb — “decide” becomes “make a decision.”
• It’s not grammatically wrong, but stacking these phrases makes sentences longer, flatter, and more abstract than they need to be.
• The fix is mechanical: find the buried verb inside the noun and let it do the work again.
• Heavy nominalization is a quiet AI tell because models lean on safe, padded noun phrases — converting them back to verbs adds the rhythm and directness detectors read as human.
• You don’t need to kill every noun. The goal is fewer dead verbs, not a war on the suffix -tion.

What is nominalization in writing?

Nominalization is the act of turning a verb or an adjective into a noun. “Decide” becomes “decision.” “Fail” becomes “failure.” “Happy” becomes “happiness.” That part is harmless — English does this constantly, and we need those nouns.

The trouble starts when the noun replaces the verb in your actual sentence. Once the action is locked inside a noun, the sentence needs a new verb to function, and that new verb is almost always something limp: *make*, *give*, *have*, *conduct*, *perform*, *provide*. So “I decided” swells into “I made a decision.” Nothing is wrong with it. It’s just bigger, slower, and one step further from the thing you mean.

Writer Helen Sword nicknamed these “zombie nouns” in a 2012 New York Times column, because they shuffle around the sentence eating the life out of verbs. The label sticks because it captures the feeling: the words are technically alive, but the energy is gone.

A quick anatomy

Look at the structure of a classic nominalization:

> conduct an analysis of the data

The real verb you want is *analyze*. But it’s been frozen into the noun *analysis*, so the sentence has to draft in *conduct* just to have something to do. Strip it back:

> analyze the data

Three words instead of five, and the verb finally points at the thing it’s acting on. That’s the whole move.

Why it makes writing feel flat

Verbs carry motion. When you bury the verb in a noun, you trade a moving sentence for a static one. A few of these in a paragraph and the prose starts to feel like it’s describing actions from a great distance rather than performing them.

There’s also a length tax. Every nominalization needs a helper verb and usually a preposition or two (*of*, *regarding*, *with respect to*). Those words pile up. A page written mostly in nominalizations runs longer than the same page written in verbs, and the reader has to do more work to extract less meaning.

And here’s the sneaky part: nominalization lets you hide who’s doing what. “A decision was made” tells you nothing about who decided. Bureaucratic and corporate writing loves this exact dodge, which is part of why heavily nominalized text reads as evasive even when no one’s hiding anything.

The AI-detection angle

Large language models are fluent at this pattern, maybe too fluent. Trained on mountains of formal, edited prose — academic papers, reports, documentation — they reach for safe, padded noun phrases by default. Ask a model to explain something and you’ll often get “the utilization of this approach enables the optimization of…” instead of “this approach optimizes…”

That matters because AI detectors don’t read for meaning. They read for statistical fingerprints: how predictable each word is, and how much your sentence lengths vary. Text that leans on the same noun-heavy scaffolding over and over comes out uniform and predictable — and that uniformity is exactly what tends to raise a detector’s score. If you want the longer explanation of why uniform, predictable text reads as machine-made, it’s worth a read.

Converting nominalizations back to verbs does two useful things at once. It shortens some sentences, which breaks up the even rhythm. And it forces concrete, active phrasing, which is harder for a model to generate by default. You’re not gaming anything — you’re just editing the way a sharp human editor would, and that’s the point.

How to fix a nominalization, step by step

The repair is almost mechanical once you know the pattern.

1. Find the weak verb. Scan for *make, give, have, take, conduct, perform, provide, reach, carry out*. These verbs rarely do real work. Where you find one, the action is usually hiding nearby.

2. Find the buried verb. Look at the noun sitting next to that weak verb. Did it used to be a verb? *Decision → decide. Analysis → analyze. Assumption → assume. Discussion → discuss. Recommendation → recommend.*

3. Promote the buried verb and delete the scaffolding. Drop the weak verb and the extra prepositions, and let the recovered verb carry the sentence.

Here’s a before-and-after run:

Nominalized	Fixed
We will *make a decision* by Friday.	We will *decide* by Friday.
She *gave consideration to* the offer.	She *considered* the offer.
They *conducted an investigation into* the leak.	They *investigated* the leak.
The report *provides a description of* the process.	The report *describes* the process.
Let’s *have a discussion about* the budget.	Let’s *discuss* the budget.

Notice the fixed column isn’t just shorter — it tells you who acts and what they do, with no fog in between.

A mini-scenario

Say a student is editing a discussion-board post and it opens like this:

> “In this post, I will make an examination of the author’s argument and provide an evaluation of its effectiveness through a consideration of the evidence presented.”

Twenty-six words, four nominalizations (*examination, evaluation, consideration*, plus the wind-up “in this post”), and it hasn’t said anything yet. Recover the verbs:

> “I’ll examine the author’s argument and weigh how well the evidence supports it.”

Twelve words. Same claim, but it moves. It also sounds like a person thinking, not a template filling itself out. That second version is the kind of sentence that survives a read-aloud test — and a detector pass.

What *not* to do

Don’t go hunting every noun that ends in *-tion* or *-ment* like it owes you money. Some nominalizations are the cleanest option available. “The investigation lasted three months” is fine; rewriting it to keep “investigate” as a verb would be clumsier, not better. “Information,” “equipment,” and “department” aren’t disguised verbs at all — they’re just nouns.

Don’t shred every sentence into four-word fragments either. The goal is to recover wasted energy, not to make everything terse. Once you’ve freed up the words a nominalization was hogging, spend them on something useful: a concrete detail, an example, a longer sentence that earns its length. Mixing short and long is what gives prose its pulse.

And don’t treat this as a substitute for actually saying something. Tightening “make a decision” to “decide” won’t rescue a paragraph that’s vague at its core. It just stops you from hiding the vagueness behind extra words.

A 60-second self-edit pass

Before you submit anything, try this:

• Search your draft for *make*, *conduct*, *provide*, *perform*, and *consideration*. Each hit is a candidate.
• Read the piece out loud. Sentences that feel like wading through mud are usually nominalization-heavy.
• For your three slowest sentences, find the buried verb and set it free.

That’s enough to lift the whole piece a level. You’ll find more writing-craft guides on the blog if you want to keep sharpening — nominalization is just one of several habits that flatten otherwise good writing.

Frequently asked questions

Is nominalization grammatically incorrect? No. Every nominalized phrase here is perfectly grammatical, and sometimes a noun is the right call. The problem is volume and habit. When most of your sentences route action through a noun plus a filler verb, the writing turns abstract and slow even though no single sentence is wrong. It’s a style issue, not an error.

Are all words ending in -tion or -ment bad? Not at all. “Information,” “government,” and “education” are real things, not disguised verbs. A nominalization only causes trouble when it hides an action you could state more directly. “The committee reached a decision” hides “decided.” “The government raised taxes” hides nothing. Look for the buried verb first.

Does fixing nominalizations actually help with AI detection? Indirectly, yes. Detectors respond to predictability and uniform rhythm, and models overuse padded noun phrases, so a draft full of them reads flat. Converting those phrases to verbs shortens sentences, varies their length, and adds directness — the same qualities a human editor pushes for. It’s not a magic switch, just good editing that happens to move the needle.

How do I spot one quickly? Scan for “weak verb + noun”: make a decision, give consideration, conduct an analysis, provide assistance, perform an evaluation. The verbs are doing almost nothing while the real action sits in the noun. When you see that combo, ask whether the noun used to be a verb. Usually it did.

Won’t shorter sentences make my writing choppy? Only if every sentence ends up short. The point is to free up the energy padding was wasting, then spend those words where they help. A mix of short and long beats a wall of even, noun-heavy lines.

The bottom line

Nominalization is one of the easiest stylistic fixes you’ll ever learn, and one of the highest-leverage. Recover your verbs and your writing gets shorter, clearer, and noticeably more human — to readers and to detectors alike. When you’ve tightened a draft and want to see how it scores, run it through PaperBleach and check the heatmap before you hit submit.