"Comprehensible input" gets thrown around a lot in language-learning circles, usually as a buzzword attached to whatever app or method someone is selling. The term actually comes from a specific place: the work of linguist Stephen Krashen, starting in the late 1970s and laid out fully in his 1982 book "Principles and Practice in Second Language Acquisition." It's worth knowing what the theory actually claims, because it explains something a lot of adult Spanish learners run into and can't quite name: why years of lessons and apps can leave someone able to conjugate every verb and still unable to follow a normal conversation.
The five ideas behind one claim
Krashen's framework, often called the Monitor Model, is built from five linked hypotheses: the acquisition-learning distinction, the natural order hypothesis, the monitor hypothesis, the input hypothesis, and the affective filter hypothesis. Two of the five matter most for a Spanish learner trying to figure out why their progress has stalled.
The first is the acquisition-learning distinction: Krashen argued that "acquiring" a language, the way a child picks up their first language, and "learning" a language, the conscious study of its rules, are separate processes that don't feed into each other the way people assume. You can learn the rule for when to use the subjunctive and still not acquire the ability to use it correctly without thinking. The natural order hypothesis adds that grammatical structures tend to get acquired in a fairly predictable sequence regardless of the order a textbook teaches them in, which is part of why drilling a rule out of sequence doesn't necessarily speed anything up. The second hypothesis that matters most day to day is the input hypothesis itself, the piece that gets cited constantly and is usually what people mean when they say "comprehensible input."
The input hypothesis and the i+1 idea
The input hypothesis states that people acquire language by understanding messages that are slightly beyond their current level, a point Krashen labeled i+1, where "i" is a learner's current stage and "+1" is the next stretch just past it. Understand enough of a sentence to follow the meaning, and the parts you don't yet know get filled in by context, repetition, and exposure over time. Give a learner content that's entirely at their level, and there's nothing left to acquire. Give them content that's far beyond it, and there's nothing to hold onto.
This is a specific, falsifiable claim, not a vague suggestion to "read more." It's also why gap-fill exercises and isolated vocabulary drills sit outside the theory almost entirely: a fill-in-the-blank exercise tests whether you already know a form, in a sentence built for testing, not whether you can extract new meaning from language you haven't fully mastered yet. The input hypothesis is about the second thing.
The affective filter: why stress gets in the way
The fifth hypothesis in Krashen's model, the affective filter hypothesis, is less quoted but explains something most language learners have felt directly. It proposes that anxiety, low motivation, or a lack of confidence raise a mental "filter" that blocks input from being processed for acquisition, even when the input itself is understandable. A learner who is tense about being corrected, or who feels judged for making mistakes out loud, can be looking at exactly the right material and still not absorb much of it.
This is part of why classroom settings, cold-call questions, timed quizzes, can work against the very acquisition they're meant to produce. It's also an argument for private, low-stakes reading as a starting point rather than the only argument for skipping conversation practice altogether. The two aren't in competition; a learner who reads and listens under low pressure builds the base that later makes real conversation less nerve-racking.
What critics actually say
Krashen's model isn't universally accepted, and being direct about that matters more than pretending a 1982 theory settled the question. Critics have pointed out that the input hypothesis is hard to test rigorously, since "comprehensible" is difficult to define precisely and i+1 can't be measured with much accuracy for an individual learner. A 2025 article in Frontiers in Psychology argued for a more neuro-ecological view of language learning that treats Krashen's framework as incomplete on its own, particularly around how much explicit instruction and interaction contribute alongside input.
None of that erases the core, more modest claim that holds up across decades of applied linguistics research since: understanding real language in context builds ability in a way that memorizing isolated rules and forms does not, on its own, reliably do. The debate is mostly about how much input matters relative to other factors, output practice, explicit grammar instruction, social interaction, not about whether it matters at all.
What this means for how you actually practice
Taken together, the acquisition-learning distinction, the input hypothesis, and the affective filter hypothesis point toward a specific, practical shift for an adult Spanish learner: spend more time reading and listening to real Spanish that you can mostly follow, in a setting with no pressure to perform, and treat vocabulary drills and grammar study as support for that activity rather than a replacement for it.
That's a different sequence than the one most courses and apps default to, which usually front-load isolated drills and treat reading real material as something you graduate into once you're "ready." The theory suggests the order should run closer to the opposite: get into real, understandable Spanish as early as the i+1 level allows, and let drilling fill in specific gaps you actually run into.
In practice, that looks less like finishing a grammar unit before touching a short story, and more like picking a story graded close to your level on day one, reading it with almost full comprehension, and only pausing to look up the handful of words or forms that actually block meaning. The subjunctive, to return to that example, gets acquired faster by running into it dozens of times in sentences you mostly already understand than by memorizing the conjugation table in isolation and hoping it transfers.
Where habloHack fits
HabloHack is built around that sequence rather than the drill-first one. Stories are graded across the CEFR scale from A1 to C2, so a learner can start close to their own i+1 zone instead of guessing at native-speed material or staying stuck on textbook sentences built for testing grammar rather than for reading. When a word breaks the flow, tapping it gives the meaning, pronunciation, and an example sentence without leaving the page, which keeps the reading itself, the actual comprehensible input, moving instead of stalling out at a dictionary lookup.
The words a learner looks up or chooses to save build personal, official, and community vocabulary lists, and reviewing them runs on FSRS, a spaced repetition algorithm that calculates when each specific word is about to be forgotten rather than showing every card on a fixed schedule. None of that replaces reading. It's built to support it, which is closer to what the theory this article is about actually argues for.