I wanted to share a moment that truly changed how I design activities.

I was at the CI Summit in Savannah, sitting in one of Eric Herman’s sessions, feeling pretty good about my comprehensible input game. My students understood what I said. They could follow a story, answer questions, and read a text. I was doing my best to provide compelling, comprehensible input. And then Eric said something that I haven’t stopped thinking about:

Students can comprehend input without actually processing it.

Understanding and processing are not the same thing.

I sat with that for a while. Because if it’s true, and it is, then it means a student can walk out of my class having understood everything I said and still not have made the form-meaning connections that lead to acquisition. They heard me. They got the gist. They did not necessarily notice that the verb ending changed, or that the word order was doing grammatical work, or that the plural marker was even there.

This post is about why that happens, and what the research says we can do about it.

NOTE: Eric does great work in sharing research. Check out his site!

The Gaslight: “If They Understood It, They Processed It”

This is the assumption that sneaks into many CI classrooms, including mine for a while: if students can understand the message, the form will eventually sort itself out through exposure. Comprehensible input is enough.

Here’s the thing. Comprehensible input IS necessary. But “necessary” and “sufficient” are different words, and it’s worth pausing on what “comprehensible” even means here. Terry Waltz has made a really useful distinction between comprehensible input (input that could theoretically be understood) and comprehended input (input that a learner actually understood). That’s already an important upgrade. But even comprehended input is not the same as processed input. The research on Input Processing makes a strong case that comprehension and acquisition are not the same pipeline.

So we’re actually looking at a three-step ladder: comprehensible, then comprehended, then processed. Each step is a real upgrade, and each one has implications for how you design instruction.

The Gatekeep: What Input Processing Actually Says

VanPatten’s model centers on a key concept: intake.

Intake is the part of input that actually gets processed. It’s the moment when a form-meaning connection is made, when a learner doesn’t just encounter a form but registers it, assigns a purpose to it, and starts to incorporate it into their developing linguistic system. Input can wash over a learner without ever becoming intake. Intake is the goal.

There are two core principles that explain when intake is more or less likely to happen.

Principle 1: Primacy of Meaning

Learners process input for meaning before they process it for form.

This is not bad news. It’s just how human language processing works. When we encounter language, we go for the message first. The grammatical machinery tends to come second, if at all.

For comprehension-based teachers, the first principle itself probably isn’t surprising. We already know that meaning comes first. The interesting stuff is in the sub-principles, because they tell us which parts of the meaning-grabbing process tend to crowd out form.

The Lexical Preference Principle: If learners can get grammatical information from vocabulary, they won’t pay attention to form. This one has real classroom implications. If you start every sentence about the past with yesterday or last week, you’ve removed any reason for a student to notice the verb ending. The lexical item is doing all the grammatical work, and the form becomes invisible. Not because students are lazy, but because they’re efficient.

The Sentence Location Principle: Where a form appears in a sentence affects how likely students are to process it. Items at the beginning of an utterance are most accessible; items buried in the middle are hardest to process. Farley illustrates this with the sentence John hates movies. The third-person -s on hates is mid-sentence, which makes it the least likely to be processed. The -s on movies is at the end of the sentence. John is the very first word, and the most likely to be processed. Same suffix, very different processing likelihood based on position alone.

The Preference for Nonredundancy: When a sentence contains multiple cues to the same grammatical meaning, learners will process the form that does the most unique work. If last weekend already signals past tense, the verb ending isn’t adding new information. It’s redundant, and students will skip it. Design input where the form is the only cue to meaning, and you give students a reason to actually notice it.

The Primacy of Content Words Principle: Grammatical function words, like articles, prepositions, and agreement markers, tend to get glossed over because they feel like filler compared to content words. This is part of why gender agreement in Spanish is so persistently late to develop. La versus el rarely changes the meaning of a sentence in any way that matters for comprehension, so learners accurately judge that they can ignore it, at least for now.

The Meaning Before Nonmeaning Principle: Forms that carry no semantic weight are harder to acquire because learners have no processing reason to attend to them. Noun-adjective agreement in Spanish is a classic example. The -a ending on an adjective is grammatically meaningful, but it rarely changes what the sentence means. That’s a processing obstacle you’ll need to design around.

The Availability of Resources Principle: You’ve heard shelter vocabulary, not grammar. This principle is the research backing for that idea. If learners are using all their cognitive resources just to figure out what the words mean, they have nothing left over for processing form. Flooding input with unfamiliar vocabulary doesn’t just make comprehension harder. It makes form processing nearly impossible.

Principle 2: The First Noun Principle

Learners will assume that the first noun or pronoun in a sentence is the subject, the agent, the one doing the action.

This holds across language types, including Object-Verb-Subject languages, and it’s not simply L1 transfer. It’s a default processing strategy. In most sentences, it works perfectly. But it creates predictable errors with passive constructions, sentences that start with an object, and any structure where the first noun is not the agent.

John is adored by his father often gets processed as John adores his father, because John came first, and as we know, the first noun is the doer!

BUT, three sub-principles can help sort this out:

The Lexical Semantics Principle: Swap out John for golf, and suddenly the first noun can’t logically be an agent. Golf adores his father is nonsense, so learners override the default and process correctly. Inanimate or logically implausible first nouns reduce the pull of the first-noun strategy.

The Event Probability Principle: Similarly, if the most likely real-world interpretation doesn’t match a first-noun-as-agent reading, learners are more likely to get it right. The parrot was adored by John’s father is easier to process correctly than John was adored by the parrot, because parrots adoring people is unlikely enough for learners to take another look.

The Contextual Constraint Principle: If learners already have context from a story, a prior scene, or shared classroom knowledge, they can override the first noun strategy because they already know who’s doing what to whom. This is one of the reasons story-based instruction supports form processing. The narrative gives learners the scaffolding they need.

The Girlboss: What This Means for Your Classroom

Here’s what I took from all of this, put into practical terms.

Your input design is an intervention. Every time you decide where to put a form in a sentence, whether to include a lexical time marker, or how much new vocabulary to front-load, you’re making a processing decision. The research gives you a framework for making those decisions more intentionally.

Comprehensible input is the floor, not the ceiling. Students’ understanding of the message is necessary, but it’s not enough if you want them to process the forms embedded in it. Meaningful, comprehensible input is where you start. But the form has to be doing visible work in the sentence for intake to happen.

You can’t just expose students to a form and wait. If every sentence that uses the preterite also starts with ayer or la semana pasada, you may be providing tons of input without ever giving students a chance to focus on the verb ending. The input is comprehended. It may not be producing intake for that form.

Start small. None of this requires an overhaul. Pick one form you think most students are ready to acquire. Look at the activities you already have and ask: Am I accidentally making this form redundant or invisible? If so, what’s one thing I could change? That’s it. That’s the start.

Next up: Part 2, How to Design Activities That Actually Force Form-Meaning Connections. That’s where Structured Input activities come in, and where we’ll get into referential vs. affective activities, how to sequence them, and what this actually looks like with real classroom content.

Further Reading:
Structured Input: Grammar Instruction for the Acquisition-Oriented Classroom- Andrew P. Farley
Proficiency-Based Instruction: Teaching Grammar for Proficiency- Catherine Ritz and Mike Travers, ACTFL
Eric Herman’s Acquisition Classroom Memo
While We’re on the Topic- Bill VanPatten, ACTFL
Common Ground- Florencia Henshaw and Maris Hawkins

Your Students Are Comprehending, but Are they…