Exploring the Four Corners of the Text Doctrine of Close Reading in AI-Integrated Classrooms

Dec 22, 2024

The machinery of the “four corners doctrine” of text as it has played out in American middle and high school Common Core literacy pedagogy since 2012 is built from a metaphor stuck in the space between extraction and disputation. Even when a plot of land is bounded by a circle, the plot must have four corners to anchor its interior and separate it from adjacent plots. Self-contained in a quadrilateral, the plot of land is deeded, titled, and recorded. Circled properties must be carefully scribed and bordered to protect every inch of ground.

The Common Core State Standards for English Language Arts explicitly codify this bounded approach to textual interpretation. Standard CCSS.ELA-LITERACY.RI.9-10.1 requires students to "cite strong and thorough textual evidence to support analysis of what the text says explicitly as well as inferences drawn from the text." The emphasis on "textual evidence" reinforces the four corners doctrine by demanding that interpretations be anchored within the text itself. Similarly, CCSS.ELA-LITERACY.RL.11-12.1 asks students to cite evidence to support both explicit meanings and inferences, but notably adds "where the text leaves matters uncertain" - implying that even uncertainty must be mapped within textual boundaries1.

This doctrinal approach is lodged in common classroom practices. Teachers routinely instruct students to "stay within the text" when writing analytical essays. Phrases like "the text states," "the author shows," and "according to the passage" become ritualistic markers of legitimate interpretation. The popular claim-evidence-reasoning (CER) framework, widely adopted under Common Core, further institutionalizes this containment of meaning by requiring students to draw their evidence exclusively from within the text's boundaries2.

Consider how this plays out in a typical high school English classroom: When analyzing Steinbeck's "Of Mice and Men," students are often discouraged from bringing in historical context about the Great Depression unless it's explicitly referenced in the text. Their analysis of Lennie's character must be built solely from Steinbeck's words, not from modern understanding of developmental disabilities. This artificial boundary-setting, while intended to build close reading skills, often creates what composition theorist David Bartholomae3 calls "the conditions of lockdown" where meaning is imprisoned within textual walls.

To speak of burglarizing a text exposes the darkest threat to the four corners doctrine. Readers-as-burglars approach the text's perimeter, testing for points of entry where language runs thin, where meaning is unguarded. Burglars listen for echoes in seemingly empty passages and search for hairline cracks between official interpretations well-known and held by local authorities. The idea is to pilfer anything—a word, a phrase, an implication—not tied, glued, nailed, nor weighted down for repurposing or retailing or keeping in one’s own plot of ground.

Like any experienced intruder, good burglars know the house of meaning has secrets. Every plot above ground has an underground, a locked cellar, buried seeds of weeds and rushes. Forbidden connections swept under adjectives, valuables swaddled in subtext, escape routes through tunnels of parentheses—burglars turn over every rock looking for special value.

The burglar metaphor reveals more than just pedagogical tensions. It mirrors fundamental questions in intellectual property law. Just as classroom burglars slip between authorized readings, legal systems have long struggled to define and police the boundaries of textual ownership. The four corners doctrine that shapes classroom interpretation finds its legal counterpart in copyright law, where the rules of intellectual property are staked out with a bit less precision than those governing land surveys

Roy Orbison had no idea that his pretty woman had tucked secret messages in her handbag to smuggle into the future, secrets that would hold Roy accountable in living sarcasm for his crudeness and mysogyny. Musicians in the future would steal from his text to humiliate him. The thievery, if thievery it is, lefts some traces, and in Roy’s case led all the way to the Supreme Court. The original text remained intact while its borrowed significance was amplified and sold on public airwaves. Were property rights violated?

If meaning can be stolen, then it must be property; if readers can trespass, then texts must be territory. Yet unlike the thief who carries away physical goods and often leaves something unheeded behind, the textual burglar very often absconds without a trace, no footprints, no fingerprints, often getting away without penalty. This paradox haunts the doctrine especially in classrooms: How can teachers police the boundaries of interpretation when every reading, even a lawful one, takes something away that belongs to the author?

The tension between containing and liberating textual meaning that we see in Common Core pedagogy mirrors a longer historical struggle in copyright law. Just as teachers wrestle with policing the boundaries of student interpretation, courts have long grappled with defining the limits of textual ownership.

Copyright Law

The four corners doctrine sits at the heart of copyright law, where the rules of intellectual property are staked out with a bit less precision than those governing land surveys. Yet unlike physical theft, the appropriation of textual meaning leaves the original meaning intact. If classroom burglars face detention, their legal counterparts face litigation. The history of copyright law is, in many ways, a history of authorized and unauthorized entry into texts.

Almost two centuries after the invention of the printing press, the issue of ownership of meaning came to a head in England and has continued uninterrupted into the 21st century. From the Statute of Anne (17104) through Google v. Oracle (20215), courts have wrestled with a unique paradox: Intellectual theft multiplies value for the public even as it depletes value for the originator.

England's Statute of Anne (1710) revolutionized textual ownership by establishing 14-year copyrights, renewable once, and by replacing monopolies favoring elite printers with author rights. When copyright terms expired, works entered public domain, an innovative and radical idea balancing authorial rewards with the free flow of knowledge. This 1710 law marked a shift from treating texts as perpetual property of the author to viewing them as temporary land grants, setting a precedent that would influence copyright law from the U.S. Constitution to modern intellectual property frameworks worldwide.

The U.S. Constitution's Copyright Clause (1787) and first Copyright Act (1790) mirrored Britain's balance of 14-year protections for creators followed by unfettered freedom for all to promote human knowledge development. Progress required assimilating both individual profit and public access. The Copyright Clause, found in Article I, Section 8, Clause 8 of the U.S. Constitution, grants Congress the power to “promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.6”

While early copyright law focused on printed texts with clear physical boundaries, modern cases like Google v. Oracle (2021) push us to reconsider what constitutes a 'text' in the digital age. This landmark case challenged traditional notions of textual boundaries by questioning whether computer code deserves the same protections as other written works.

The Google vs. Oracle (2021) dispute pushed the limits of copyright law by calling into question what we mean by “text.” Google lifted 11,500 lines of code Oracle had written to drive Oracle’s Java API. For a user an API (Application Programming Interface) is like a restaurant menu with strict rules. It lists exactly what services are available and how to order them. Just as a menu shows what dishes you can get but hides the kitchen's chaos, Java API shows what functions you can use through Oracle but hides the complex code behind them. Google took 11,500 lines of Oracle’s code from behind the menu and started using it for their own restaurant menu.

In a 6–2 ruling delivered by Justice Stephen Breyer, the Supreme Court assumed for argument’s sake that the API code was copyrightable like language in a text but focused solely on whether Google’s use was “fair use.” The Court concluded that Google’s actions qualified as fair use under U.S. copyright law7. Whether code behind an API has meaning like a traditional text or some categorically different meaning, Google used it fairly. The court left open the question of “what is a text in the context of copyright,” a question highly relevant now that generative AI is available, for a future court.

Google v. Oracle (2021) boiled down to this: Can you copy someone's API code to build something new? Google took 11,500 lines of Java's code to help app developers use Google’s Android mobile operating system. The Supreme Court said yes, Google, app developers can use this Oracle code to make new Android apps, and it’s fair for you to provide it to them. It’s fair use because Google transformed Java's API into a mobile platform, copied mostly functional (not creative) elements, took only what they needed, and didn't harm Oracle's market.

Think of it like using a familiar recipe format to write a completely different cookbook. The format helps chefs understand your recipes, but you're creating new content. Justices Alito and Thomas disagreed, saying Google should have built their own code from scratch or paid Oracle, but the majority ruled. Sometimes you can use someone else’s blueprint to build something different from it.

The Supreme Court's emphasis on transformation in Google v. Oracle opens up intriguing possibilities for how we think about fair use in the classroom in the age of artificial intelligence. If Google can legitimately steal Java's familiar syntax to help them transform resources for their users, what implications does this have for students using AI to help them think better?

Fair Use

If Google can copy 11,500 lines of Oracle's code verbatim to help developers do creative and productive work more efficiently and effectively, why can't students use AI to scaffold their thinking? Both cases involve borrowing existing structures to create something new.

The application of fair use doctrine8 to bot-assisted writing reveals intriguing parallels to the Google v. Oracle case while highlighting unique aspects of AI-generated content. When examining the first factor of fair use, purpose and character, bot-assisted writing demonstrates strong transformative qualities similar to Google's API implementation. Rather than merely reproducing existing work, the bot helps create new content, particularly in educational contexts where courts have traditionally favored fair use claims. However, commercial applications might weaken this position.

The nature of copyrighted work presents an interesting twist in the AI context. While courts typically grant stronger protection to creative versus factual works, this distinction becomes less relevant with bot-assisted writing. The bot's training encompasses both types of works, but its output is so transformed from any individual source that the original nature of the works becomes almost immaterial to the analysis. There literally is no “four corners of a text” to become the target of burglars.

Regarding amount and substantiality, bot-assisted writing presents a stark contrast to Google's targeted use of specific API code. Rather than copying substantial portions of any single work, the bot operates on patterns and possibilities derived from its training data. No individual work is substantially copied, which, like the point under amount and substantiality, strongly supports a fair use finding under this factor.

The market effect analysis also differs significantly from the Google/Oracle scenario. Where that case presented clear market impacts to consider, bot-assisted writing's effect is more diffuse, as it doesn't directly compete with or substitute for specific copyrighted works. However, the widespread adoption of AI writing assistance could impact professional writing services, introducing broader market considerations that courts will need to address.

This analysis suggests that bot-assisted writing may present an even stronger case for fair use than Google's ultimately successful defense of API implementation, particularly in educational contexts. The highly transformative nature of the use, combined with the absence of substantial copying from any single work, provides compelling support for fair use protection. Google didn't pass off Java as its own invention. It used familiar syntax to help developers build in a new space. Similarly, students using LLMs to outline essays aren't necessarily plagiarizing. They are using AI as intellectual scaffolding to construct their own arguments.

The Supreme Court emphasized that Google's borrowing served innovation by letting developers apply existing knowledge in new ways. Replace "developers" with "students" and "Java" with "writing structures," and the parallel crystallizes. Both uses transform existing tools into new creative acts. Just as Google didn't simply replicate Java but built something different with its familiar pieces, students using AI outlines can transform those frameworks into original thought.

The key lies in transformation, not in slavish copying. The question isn't whether tools were borrowed, but how they were used to transform old into new. The relationship between transformation and originality that emerges in discussions of AI-assisted writing echoes an earlier landmark case about creative reinterpretation. The case of Roy Orbison's “Pretty Woman” illuminates how textual burglary operates in both classroom and courtroom. While English teachers might question whether feminist reinterpretations of the song stay “within the text,” copyright law faced similar questions when 2 Live Crew transformed the original. This tension between containing and liberating textual meaning that we see in Common Core pedagogy mirrors a longer historical struggle in copyright law.

2 Live Crew altered the original song in several significant ways to create a parody. While the parody began with the same opening line as Orbison's original, it quickly diverged into comical and shocking lyrics. The new lyrics satirized the original by juxtaposing its romantic tone with crude and humorous commentary, aiming to critique both the song and societal norms. 2 Live Crew retained the distinctive opening bass riff of the song, but they added elements typical of hip-hop, such as "scratching" and rhythmic beats, giving the song a completely different style and tone. The group used the original song's "heart"—its most recognizable elements—to "conjure up" the original for parody purposes. Justice Souter noted that parody inherently targets the core of a work to make its commentary effective, which justified their use of these key elements9.

The parody critiqued the sentimentality and naiveté of Orbison's original song, contrasting it with a more raw depiction of street life and societal issues. This transformative intent was central to the Court's finding that the work was not merely derivative but instead offered new expression and meaning. The path from "Oh, Pretty Woman" to Java code reveals how transformation defends potential burglars against claims of theft. When 2 Live Crew turned Roy Orbison's love song into biting social commentary, they transformed not just melody but meaning.

The Supreme Court recognized that parody creates new value rather than merely appropriating the old. Similarly, when Google borrowed Java's familiar syntax, they weren't copying Oracle's intellectual property. They were transforming a desktop programming language into a mobile development platform. Both Google vs. Oracle and the Pretty Woman case pivot on a crucial distinction. Transformation that builds something new versus copying that depletes something old already owned by another is a legal principle. 2 Live Crew didn't diminish Orbison's original. They amplified its cultural resonance by giving it new context. Google didn't reduce Java's value. They expanded its utility by creating new creative space.

When T.S. Eliot declared "good poets borrow, great poets steal," he illuminated shades of gray in this peculiar crime by positioning the burglar of poems who commits theft as the “great” poet, the poet who trespasses while leaving footprints everywhere. Ought we condemn William Faulkner for his “great steal” from Shakespeare when he titled his novel “The Sound and the Fury”? As AI turns every reader and writer into a potential thief, the paradox intensifies10. How do we surveil and police the four corners of texts in an artificial dimension where trespass doesn’t involve incursion across text boundaries—the four corners are safe. AI weaves together generic textual and linguistic features that provide a computational generator with the necessary data to predict what words would make sense in this context. LLMs quote from specific, single texts with four corners only when these texts are loaded in their entirety into the machine.

These legal precedents establishing the difference between transformative use and mere copying provide a framework for rethinking how we approach student engagement with texts in the AI era. Teachers now face similar challenges to those the courts have wrestled with: How do we distinguish between AI-assisted analysis that transforms understanding and AI-generated content that merely reproduces existing ideas?

When Doctrines Break Down

When the Common Core’s four corners doctrine meets artificial intelligence, we confront a paradox about meaning, ownership, and transformation. Just as Google's appropriation of Java code and 2 Live Crew's reinvention of "Pretty Woman" demonstrated how creative transformation differs from mere theft, AI forces us to reconsider what it means to stay "within the text." The metaphorical walls we've built around texts, whether through copyright law, classroom pedagogy, or interpretive Common Core doctrine, are being reshaped by technology that doesn't so much break in as flow through.

The evolution from the Statute of Anne to Google v. Oracle shows us that intellectual property law has always struggled to balance protection with progress. Each major case has forced us to refine our understanding of what constitutes fair use and transformation. Now AI presents us with an even more complex challenge: a tool that doesn't steal from specific texts but rather synthesizes patterns across all texts. Like the textual burglar on steroids who leaves no fingerprints, AI moves through our carefully bounded territories of meaning without disturbing their corners.

This technological reality demands that we evolve beyond simplistic notions of textual trespass. Just as courts have recognized that transformation can create new value without depleting the original, educators must reconsider whether AI-assisted learning represents theft or transformation. The four corners doctrine, born in an age of printed pages and clear boundaries, must adapt to a world where meaning flows more fluidly between texts, tools, and brains.

Perhaps what we need is not stronger walls around our texts, but better understanding of how meaning transforms as it moves. When students use AI as intellectual scaffolding, they aren't breaking into the house of meaning. They're potentially learning to build new houses using time-tested blueprints. The question isn't whether they've stayed within the four corners, but whether they've transformed what they've found there into something meaningfully their own.

In this new landscape, we might do well to remember T.S. Eliot's distinction between borrowing and stealing. The great writers, like the most effective learners, don't simply borrow and copy what they find; they steal it,, transform it, make it their own. They own it. AI gives us new tools for this transformation, challenging us to redefine what constitutes authentic engagement with texts. The four corners still matter, not as prison walls that confine meaning, but as anchor points from which new understanding can grow. AI could actually turn out to reinforce the importance of the four corners doctrine by steering learners toward highly valued texts for close reading that might otherwise go unnoticed and unheeded.

Learning to Read, Reading to Learn

Discussion about this post