The Good Word

The Secret Rules of Adjective Order

A long fascinating article—or is it a fascinating long article?

Is this a big brown dog or a brown big dog? 

Photo illustration by Slate. Photo by Silense/iStock/Thinkstock

It is a lovely warm August day outside, and I am wearing a green loose top. Does the second part of that sentence sound strange to you? Perhaps you think I should have written “loose green top.” You’re not wrong (though not entirely right, because descriptivist linguistics): An intuitive code governs the way English speakers order adjectives. The rules come so naturally to us that we rarely learn about them in school, but over the past few decades language nerds have been monitoring modifiers, grouping them into categories, and straining to find logic in how people instinctively rank those categories.

If you’re someone whose reflexes scatter the moment you try to lift the veil on your unconscious, this fascinating little-known field (little-known fascinating field?) will drive you nuts. On the other hand, thinking about how adjectives work may bounce you to an epistemological Zen state, wherein you can contemplate amid flutes what it means to partake of Redness and whether former child actress means something different from child former actress. Adjectives are where the elves of language both cheat and illumine reality.

Maybe I am overqualifying this article about qualifiers (or is that the point?).

Linguists have broken the adjectival landmass into several regions. They are: general opinion or quality (“exquisite,” “terrible”), specific opinion or quality (“friendly,” “dusty”), size, shape, age, color, origin, and material. Generally, modifiers from the same region can be strung together in any order. Thomas Wolfe, writing in Look Homeward, Angel of “blistered varnished wood” and “fat limp underdone bacon,” could also have said “varnished blistered wood” or “limp fat underdone bacon.” (All five examples count as “specific opinion” words.) Likewise, if the woman in “The Idea of Order at Key West” had walked along the “tragic-gestured, ever-hooded sea,” instead of the “ever-hooded, tragic-gestured sea,” probably no one would have sent the grammar cops after Wallace Stevens.

On his blog, the linguist Neal Whitman calls adjectives from the same semantic region—the ones where swapping their placement in a sentence neither sounds strange nor scrambles the meaning—correlative. Correlative adjectives often, though not always, announce themselves through commas, and they are good at modifying nouns without talking much to each other, like exes at a mutual friend’s wedding. (Commas themselves are a more complicated matter, says John T. Beavers, a professor of linguistics at the University of Texas at Austin. Because they often function as pauses, they can isolate an adjective from whatever flow might otherwise sweep it under, abetting Adjective Order Crimes.)

Whitman distinguishes between correlative pairs of modifiers and the fussier cumulative ones. These, read in succession in a sentence, accrete sense in a specific way. For example, consider the subset of adjectives called operators, which often take part in cumulative constructions. Such terms—“former,” “alleged,” “fake”—fundamentally change the meaning of whatever follows. (An “alleged” thief may not be a thief at all.) Therefore, when dealing with operators, the precise idea you want to express determines the order of adjectives, and a furniture dealer is not at liberty to oscillate between “fake Malaysian ivory”—a material masquerading as Malaysian ivory—and “Malaysian fake ivory”—a not-ivory material from Malaysia. (For more on operator adjectives, also known as non-intersective adjectives, and their role in possible adjective ordering, I mean possible role in adjective ordering, check out Alexandra Teodorescu’s 2006 paper for the 25th West Coast Conference on Formal Linguistics.)

But what about modifiers that sound good in one order and bad in another, even if they convey the same meaning both ways? Though red big barns and big red barns are semantically identical, the second kind pleases our ears more. These tricky situations—neither pure correlation nor accumulation—generally occur when you cross the border between adjectival regions, such as size and color. When that happens, an invisible code snaps into place, and the eight categories shimmy into one magistral conga line: general opinion then specific opinion then size then shape then age then color then provenance then material.

All of which can get really confusing. For one thing, it’s hard to remember. (GSSSACPM isn’t that sticky of an acronym.) Plus, the boundary between a “general” and a “specific” opinion seems thin, with words like beautiful or sweet evoking both discrete, somewhat measurable qualities and nebulous curtains of approval. A few linguists also contest the placing of shape before age, or size after opinion. (Sure, “mean little terrier” works better than “little mean terrier,” but doesn’t a large comfortable armchair sound nicer than its inverse? And what about the trump card suggested to me by the son of one Slate colleague: “BIG STINKY FART”?) 

Still, corpora studies confirm that GSSSACPM prevails in most instances of written English. In 2003, Stephanie Wulff used a computer program to comb through thousands of texts—she found that 78 percent of adjective strings followed the rule. More minutely, when Carnegie Mellon’s Enrica Rosato searched the Corpus of Contemporary American English for occurrences of “big red [noun]” versus “red big [noun],” she turned up 382 instances of “big red” to zero of “red big” 

Is there some hidden logic underpinning this arrangement, or is it arbitrary? Since GSSSACPM more or less applies to languages around the world, many linguists think we want ripe to precede yellow for a reason. “It’s possible that these elements of universal grammar clarify our thought in some way,” says Barbara Partee, a professor emeritus of linguistics and philosophy at the University of Massachusetts-Amherst. Yet when the human race tacitly decided that shape words go before color words go before origin words, it left no record of its rationale.

The Danish grammarian Otto Jespersen was first to speculate on why our adjective conventions shook out the way they did. In his seminal 1922 volume Language: Its Nature, Development and Origin, he proposed, simply, that the more specific term always falls closer to the noun. It is a seductive hypothesis, especially when you consider phrases like “nice clean metal fork” or, from Wolfe again, “cool clarion earth.” Henry James, describing Nathaniel Hawthorne’s stories as “small cold apologues, frosty and exquisite,” cradles his noun apologues between finely tuned sense words (“cold” and “frosty”), themselves hemmed in by broader judgments (“small” and “exquisite”). It is an intuitive way to hone in on the precise object of discussion, a linguistic mirroring of the mind as it focuses and then begins to pull back. “The use of multiple adjectives … is equivalent to a sequential series of restrictions placed on the set of properties for a given noun,” Rosato wrote in her 2013 paper. If so, progressing from general adjectives to precise ones makes sense.

But Jespersen’s theory only half fits with what we know about GSSSACPM. After all, the category “specific quality” comes second in the chain, ahead of more catholic groups like color and size. In 1971, psycholinguists Sam Glucksberg and Joseph H. Danks advanced a new thesis: As you get closer to the noun, you encounter adjectives that denote more innate properties. Glucksberg and Danks gave volunteers a test in which each question was a set of six sentences. The sentences had three modifiers and one noun, and only the order of the modifiers varied. Asked to rank the statements by their acceptability, the volunteers consistently rated highest the one in which the adjectives grew more “intrinsic” as they approached the noun. For example, “My sister bought a wondrous blue-green Hawaiian gecko” would earn higher marks than “My sister bought a Hawaiian blue-green wondrous gecko.” By 1985, when linguist M.A.K. Halliday published his first edition of An Introduction to Fundamental Grammar, the notion that adjectives grow “increasingly permanent as attributes” as they tend noun-ward was itself solidifying into canon.

To contemporary language scholars, this rule also seems logical. “In general, nouns pick out the type of thing we’re talking about, and adjectives describe it,” Partee told me. She observes that the modifiers most likely to sit right next to nouns are the ones most inclined to serve as nouns in different contexts: Rubber duck. Stone wall. Even in more abstract constructions—that Henry James essay on Nathaniel Hawthorne also contained the phrases “pressing moral anxiety” and “restless individual conscience”—we often see indisputable adjectives followed by nouns-roped-into-being-adjectives followed by nouns-as-nouns. And maybe that’s how it should be: The slanting afternoon light might not always slant, but it must shine in the afternoon. In physics, things have a gravitational pull, attracting other things—why not in language too?

But try to sell the intrinsicness hypothesis to those pop groups warbling about a “long cold lonely winter.” (“Long” seems like a more stable quality than “lonely,” which reflects the speaker’s value judgment. And “cold” seems more intrinsic to winter than either of them.) Or tell it to the big bad wolf, bristling at the implication that your opinion of his character is more inherent to him than his impressive dimensions. “The exceptions are so many that it is really hard to arrive at a universal law,” Beavers admits. While “these generalizations capture something important about language, some motivation for organizing it that way,” he continues, “the rules are begging to be broken.”

When that happens, it can serve a practical purpose: As Partee points out, switching up the order of adjectives allows you to redistribute emphasis. (If you wish to buy the black small purse, not the gray one, for instance, you can communicate your priorities by placing color before size). But word crimes also produce literary effects, as in a fission lab where the splitting of bonds releases bursts of energy and irradiation. Scrambling GSSSACPM helps authors achieve a sense of spontaneity, of improvising as they go. Wolfe discovers such a rhythm, a feeling-his-way quality, when he discusses his childhood recollection of “brown tired autumn earth” and a “flat moist plug of apple tobacco.” The words appear as if out of memory’s hallucinatory fog, in the haphazard order in which they occur to the speaker. They imply an authentic and unmediated relationship between the developing thought and its verbal transcription. Or consider when Yeats tells Maud Gonne, in the poem “Adam’s Curse,” that “I strove/ to love you in the old high way of love.” There’s something archaic and untenable in the reversal of “high” (specific opinion) and “old” (age), a difficulty that foreshadows the next lines’ sad capitulation: “and yet we’d grown/ as weary-hearted as that hollow moon.”

In 2007, the psychologist David Kemmerer scanned the brains of volunteers as they read aloud sentences that either followed or upended the order of adjectives. Not surprisingly, he found that reciting strings of modifiers required more cognitive effort when the strings were beaded “wrong.” That ability to make a reader pause and concentrate can be useful to writers. After all, some might view a sentence requiring greater thought to decipher as flawed, but inspiring an incremental increase in focus is, for many writers, the point of writing. On the other hand, phrases that cleave to GSSSACPM can have a lovely fluidity—in her poem “Question,” May Swenson asks: “How will I know/ in thicket ahead/ is danger or treasure/ when Body my good/ bright dog is dead?” The way Swenson’s “good bright dog” (general opinion, specific opinion, noun) trots across the stanza just as we expect him to says something about his obedience, his reliability, and how lost we’ll be when he’s gone.

While Body our good bright dog still lives, though, what does the order of adjectives mean for us? It means that we will likely ask for a nice hot cup of tea, not a hot nice cup of tea, unless perhaps a cold nice cup of tea is also available. It means that our angry red sunburn will be forever angry before it is red, our huge fuzzy blanket huge before it is fuzzy. Mostly, it is just a reminder that the invisible rules of language assert themselves everywhere, even when we break them, even when we don’t think about them, even when they don’t strike us as a great important big old deal.

See all the pieces in this month’s Slate Book Review.
Sign up for the
Slate Book Review monthly newsletter.