What the Words Say
“You’re done, Tucson,” Taco said, spittle on his lips, the crunch of gravel in his voice, teeth in a grimace.
What option best characterizes what these words say?
A. Taco is threatening Tucson.
B. Taco is finishing a conversation with Tucson.
C. Taco is telling Tucson he is fired.
D. There is not enough information to determine what the words say.
The correct answer, of course, is D, yet every reader knows it is A. My bet is you want to argue about it. But you know it’s D because you bring a life to the page. Genre, frame of reference, gestures, the cadence of menace, too much of watching how people bare their teeth. That knowledge is not a higher stage of comprehension built on a literal foundation. It is comprehension, full stop.
The multiple-choice question sounds simple enough. Words always say something. Otherwise, they would not be words, but barks or honks or snorts. Yet words rarely mean what they say, and what they say rarely captures what they mean.
What readers must do with Tucson and Taco in a classroom has a name: reading comprehension. Hidden in the folds of the American perspective on comprehension is an unflappable belief that once a child knows all the words, that child can comprehend a text.
Which words from Taco’s mouth do you not understand?
Pedagogically, we sometimes name this the instrumental hypothesis, i.e., vocabulary is instrumental to comprehension. The implication is, of course, that explicit vocabulary instruction prior to reading a text improves comprehension. By modus ponens, when you teach vocabulary, you teach comprehension.
Pumpkin
“You have to go, Pumpkin. It isn’t your time any more.”
Same grammar with even less information about the speaker. A declaration aimed at someone with a pet name. Same taxonomy box on a standardized test: How well does this child understand what the words say within the four corners of the text?
But Pumpkin isn’t Tucson. “Pumpkin” connotes tenderness the way “Tucson” connotes danger with a racist undertone being spoken to by Taco, not by virtue of what the words say, but in what life in the U.S. teaches.
“Any more” does quiet but significant work. It concedes a time before. Something is being lost, not just ended. Is this a parent at a bedside? A vet’s office? An aging coach? The words do not say what they mean. The reader, drawing on reservoirs of appointments, love and loss that no test item can reach, decides.
A Grand Bargain
“That comes to 48 thousand, two hundred, forty-three dollars and 57 cents — and that’s a bargain!”
No threat. No tenderness. No one to miss or to grieve. Every word is plain, every number legible. If words say anything, if literal comprehension exists anywhere, it should exist here.
But look what a reader does. The reader doesn’t comprehend $48,243.57. There’s nothing to say. The reader registers it and moves past it because the number in this testing context functions rhetorically, not mathematically. Its precision says: Do not argue.
A reader who stopped to actually hold forty-eight thousand two hundred forty-three dollars and fifty-seven cents in mind would be reading worse, not better, because they would be chasing what the words say at the expense of what they mean, hollow as that meaning is.
And “that’s a bargain” — is it? The exclamation mark speaks volumes. You suspect a salesman? Maybe it is. Maybe it isn’t. Maybe it is a bargain. Nothing in the words settles the question. You’ve been sold things, you’ve sold things, you can name that tune in six syllables. Still, what do you know from the sentence?
Is an exclamation mark a word? Why and why not?
Pedagogically, we find ourselves in the land of the reverse instrumental hypothesis, i.e., comprehension is instrumental to learning, using, and grasping vocabulary. How can anyone understand what the word “done” says vis a vis Tucson without comprehension—or at least provisional comprehension? By modus ponens, when you teach comprehension, you teach vocabulary.
What the Words Say
Comprehension is routinely translated in the U.S. as “understanding what the words say.” Getting the gist, having a clear understanding, seeing the full picture.
"Gist" comes from Anglo-French — cest action gist, "this action lies" — a legal term for where a case rests, its essential ground, not its details nor its words1. The place where it lies, which is a beautiful ambiguity, because meaning lies in both aspects: It rests somewhere, and it still doesn't tell the whole truth.
Gist is the work of the reader. Walter Kintsch knew this full well. The situation model of text comprehension isn't a transcription of the text; it's a gist the reader purposefully fabricates from text and world knowledge, and it's better than what the words say because it's integrated from one life into another life.
An adolescent who reads the Taco sentence and says "that guy's about to hurt somebody" has the gist, though they cannot point to a single word that says so. An adolescent reading the Taco sentence in one of my English classes at James Rutter Middle School in the 1990s, a Title I school with a sizable Hispanic population, would have had a field day right then and there.
"What the words say" treats words as artificial agents speaking on behalf of the author, as if a word opens like a downloaded app with identical affordances in any pair of eyes that happen to grace a portal. Indeed, LLMs threaten the status quo because “what words say” is often not even what words mean, not even close. Words can be hallucinated.
If comprehension were truly literal, if meaning were given rather than taken, composition would have a literal counterpart: Writing would be simply depositing meaning in texts, text would be a warehouse for words. Nobody says "literal composition" because everyone intuitively understands that writing is not mental dictation.
Not extending that understanding to reading is a significant problem for reading pedagogy. It anchors reading taxonomies, which structure reading activities from grunt work first to engagement later. Despite persistent, insistent messages from scholars about the recursive nature of reading and writing mechanisms, teachers are resistant.
Literal comprehension sits at the bottom of every hierarchy from Bloom’s six levels to schema theory’s three levels. Which, for some kids, means never reaching high ground. Maybe that’s what the classroom forgets. Students bring more meaning than any test can measure
Practice Quiz
Looking up words in a dictionary while reading supports comprehension.
A. This is consistent with the instrumental hypothesis. Vocabulary knowledge leads directly to comprehension, so a dictionary accelerates the process.
B. A dictionary provides definitions, but “done” means something different to Tucson than it does to Pumpkin, and no definition resolves the difference; the reader’s provisional comprehension does.
C. Dictionaries are useful for decoding unfamiliar words, but the essay suggests they should be supplemented with explicit instruction in genre and context clues.
D. The essay supports this claim that dictionaries are helpful because gist is built from word meanings, and a dictionary is the most efficient source of word meanings.
Literary reading differs from scientific reading in that one type draws on the imagination and the other on logic.
A. The essay supports this distinction because the Taco and Pumpkin sentences require imagination, while the $48,243.57 sentence requires logical, mathematical processing.
B. The essay undermines this distinction because the reader of the transaction sentence relies on the same life-sourced, inferential machinery, i.e., suspicion of a salesman, the rhetoric of precision, that the reader of the threat does.
C. The essay suggests that literary reading is harder than scientific reading because literary texts are more ambiguous and require more background knowledge.
D. The essay draws no distinction between literary and scientific reading because both depend on the instrumental hypothesis and explicit vocabulary instruction.
Reading and writing are two sides of a coin.
A. The essay rejects this metaphor because reading is primarily receptive while writing is primarily productive, and the two processes share little cognitive ground.
B. The essay supports this metaphor because both reading and writing depend on explicit vocabulary instruction as described by the instrumental hypothesis.
C. The essay argues that the metaphor is true but that American pedagogy honors only one side of the coin. Everyone understands that writing is construction, yet reading is still treated as extraction.
D. The essay claims that reading and writing are recursive but occupy different levels of Bloom’s taxonomy, with reading at the base and writing near the top.
Anglo-French legal phrase cest action gist “this action lies (in this matter),” from Old French gist (en) “it lies in, it consists in.” Gist here is the 3rd person singular of Old French gésir “to lie,” from Latin iacet “it lies,” from iacēre “to lie, rest.” So, the core verbal root of gist is “to lie, to rest (in something), to consist in (something),” as in “the action lies in this ground.”