Skip to content

Combine intentional and incidental vocabulary learning rather than relying on incidental learning alone

Principle · All principles
Evidence · 25 claims (16 for, 8 mixed, 1 against) · 17 studies (6 causal, 5 quant-synthesis, 5 review, 1 associational), q2–q3 · 2 of 17 report an effect size · 20 claims rest on one study

Conditional relationship

This page owns the balance between deliberate word study and learning words from meaning-focused input over a course. For a second-language learner (and, by extension, a first-language reader) who is building general vocabulary, a combination of the two modes is expected to produce more words known and better retention than either mode alone: deliberate study of selected words (meaning given, then recalled and used) supplies the words that make texts comprehensible, and reading or listening for meaning, with glosses and recurring words, then adds and deepens words that would take too long to teach one by one. The intentional share should be largest while the learner knows too few of the surrounding words to infer meaning from context, and should shrink as coverage of the learner's texts rises. The capability is word knowledge (verbal-association for form and meaning) and its use in comprehension and production; the outcomes are immediate-performance on taught and met words and delayed-retention.

Two claims test the combination itself, and both are weak tests. One second-hand report of a study of 93 Chinese university ESL students found that both incidental and intentional learning produced gains and the combined instruction produced greater gains and better retention (Combining intentional and incidental vocabulary learning produces greater gains and better retention than either mode alone [+M]). A 1986 meta-analysis of vocabulary instruction, read from its abstract, concluded that the most effective methods combined definitional and contextual information, gave more than one or two exposures and asked for deeper processing, without an effect size for that comparison (Pairing Contextual Encounters With Explicit Instruction Produces Stronger Vocabulary Outcomes Than Either Alone [+M]). One small trial with first-year ESL students added mobile listening to usual vocabulary teaching and found higher end-of-unit exam scores than usual teaching alone (Listening to vocabulary lessons while physically active improves End-of-Unit exam performance over traditional intentional learning alone [+M]), though the added component also changed time and activity. The two halves are better supported on their own: incidental learning from meaning-focused input is reliable but picks up only a small share of the unknown words met (Incidental vocabulary exposure produces reliable learning, but learners pick up only a small share of the unknown words they meet [+S]), and one second-hand direct comparison found intentional learning gave more, faster gains and better retention than incidental learning (Intentional vocabulary learning yields higher, faster gains and better retention than incidental learning through reading [+M]). No claim in the wiki tests the order of the two modes, the ratio between them, the coverage at which to shift the balance, or how many encounters a word needs; those parts of the design below are proposals. One narrative review in the wiki argues the opposite direction, that incidental acquisition is more powerful than intentional learning (Incidental vocabulary acquisition through comprehensible input is more powerful than intentional vocabulary learning [-W]); it is weighed below.

The relationship is conditional on four things: how many of the surrounding words in the learner's texts the learner already knows (inference from context needs most of them; Correct word inference in reading requires knowing roughly 95-98% of surrounding words, corresponding to about 3,000-5,000 word families [+M], second-hand); how much meaning-focused input the learner actually gets each week; how often target words recur in it; and the outcome and horizon (recognising a word's meaning, recalling it, or using it in speech and writing, now or weeks later). The evidence here is mostly from second-language learners at university; carrying it to children or to adults learning the terms of a job is an extrapolation.

How the converted sibling sits beside this page: Explicit Instruction: Vocabulary holds how to teach a selected set of words (selection, introduction, active use, spaced review, a "known" criterion). This page does not repeat that model; it holds the decision around it: how much of a learner's vocabulary time goes to deliberate study and how much to reading and listening for meaning, how the two are joined within a session, and how the balance shifts across a course. Accessible Vocabulary & Syntax holds which words to remove from a task or text so that it can be read at all.

Default design

Two claims test the combination and one small trial adds input to deliberate study; none tests the order, ratio or dose. Each step names its evidence status, and the quantities are proposals to set locally and record.

  1. Check coverage before choosing the mode. Take a 100-word sample of the next text the learner will read or hear and have the learner mark the words they do not know (or test a sample of them). If fewer than about 95 of 100 are known, do not rely on inference from context: teach or gloss more. Comfortable reading needs about 98 of 100. The 95% and 98% figures are reported second-hand from one review (Correct word inference in reading requires knowing roughly 95-98% of surrounding words… [+M]); the 100-word sampling procedure is an untested proposal.
  2. While coverage is low, lead with deliberate study of the most frequent words. A learner who does not yet know roughly the first thousand words (the earlier page's threshold) gets most vocabulary time as deliberate study: 10–15 minutes a session on 5–8 high-frequency words, with meaning, examples and recall, and spaced review on cards or an app. Input in this phase is graded or simplified text, or text with most unknown words glossed. The threshold is from the earlier page and untested here; computer-based spaced review beside classroom teaching gave much higher end-of-semester credit than classroom teaching alone in one study of EFL cadets (Supplemental computer-based spaced repetition activities nearly triple long-term vocabulary retention… [+M]).
  3. Pre-teach a few key words, then read or listen for meaning. In each session with a text: 5–8 minutes on 3–5 words the text needs and the learner will meet again, then 15–20 minutes reading or listening for meaning, with a comprehension purpose rather than a word hunt. Combining definitions with contextual encounters is supported by the two combination claims above (Pairing Contextual Encounters… [+M]; Combining intentional and incidental vocabulary learning… [+M]). Which mode comes first is untested in the wiki; the minutes are untested proposals.
  4. Gloss the other unknown words in the margin; leave a few for the dictionary. Give a short marginal gloss (first-language or simple target-language meaning, a picture for concrete words) for unknown words that matter to the text and were not pre-taught, and ask learners to look up two or three more. Glosses and dictionary use each helped incidental learning in a randomised study of 78 advanced university learners of French, reported in a review (Marginal glosses, dictionary use, and repeated occurrence of unknown words each positively affect incidental vocabulary learning [+M]); a second-hand report of a study of 125 Chinese ESL freshmen found that tasks with more search and evaluation (glosses plus dictionary) gave better retention and that marginal glosses were the more efficient aid (Tasks with higher involvement load… [+M]). Meta-analyses of glossing are reported second-hand as finding medium-to-large effects, larger for text-plus-picture glosses (Glossing yields medium-to-large positive effects on L2 vocabulary learning… [+W]).
  5. Choose or write texts in which the target words recur. A word met once is rarely learned from context; plan texts, or sets of short texts on one topic, in which each target word appears several times, ideally spread over days rather than in one passage. A default of at least five or six encounters across a week is an untested proposal. The number of encounters goes with incidental learning across second-language studies, a correlation rather than a tested dose (The more often learners meet a word while reading, the more likely they are to learn it incidentally… [+M]; Vocabulary Knowledge Grows Incrementally With Repeated Encounters In Varied Contexts [+M]).
  6. After the input, turn met words into studied words. 5–10 minutes: learners pick 3–5 words from the text they want to keep, write the sentence they met it in, recall its meaning without the gloss, and use it in a sentence of their own or a short spoken answer. This is the intentional half applied to words the input supplied. Supported in direction by the combination claims; a second-hand report found that adult ESL students who read a short story could use the new words in sentences while students who had studied dictionary meanings in isolation could not (Adult ESL students who read a short story could use incidentally acquired words in sentences… [~M]), which argues for keeping the context with the word. The routine is an untested proposal.
  7. Bring met and taught words back on a spaced schedule. Words from the input lose ground over time if nothing brings them back (Incidental vocabulary gains show retention problems, with losses over time moderated by feedback timing and topic interest [~M]). Review each kept word after about 1, 3, 7 and 14 days, recall first and meaning shown after, 5 minutes at the start of a session or on an app (an untested schedule; spaced review per step 2).
  8. Protect the volume of meaning-focused input. Because each encounter teaches a small share of the unknown words met, gains from input come from volume: plan at least 60–90 minutes a week of reading or listening for meaning at a level where coverage is about 95% or higher, in or out of class (an untested dose). Listening and reading-while-listening yielded proportions of words learned similar to reading in one meta-analysis (L2 learners pick up vocabulary incidentally from meaning-focused input… [+M]; Incidental vocabulary exposure produces reliable learning, but learners pick up only a small share… [+S]). Let learners choose topics they find interesting where the course allows: one second-hand report found growth from interesting passages lasted (Incidental vocabulary gains show retention problems… [~M]).
  9. Shift the balance across the course, and move on by coverage and recall. A default arc, untested: early (coverage of graded texts below about 95%): about two thirds of vocabulary time on deliberate study of high-frequency words, input graded and heavily glossed, the teacher supplying meanings. Middle: about half and half; fewer glosses, learners infer a meaning from context first and check it against the gloss or dictionary. Late (about 95% coverage or more of the texts the learner needs): about a third on deliberate study, kept for words input will not supply often enough (low-frequency academic or technical words) and for the after-input routine of step 6; most time is extensive reading and listening with few or no glosses. Move a learner to the next text level when a 100-word sample of it shows about 95% known and they recall the meaning of most reviewed words (say four in five) at the two-week review. The ratios and criteria are untested proposals; the direction (intentional learning first builds the words later incidental learning depends on) is from the earlier page and the coverage claim.

When a condition is not met: if the learner cannot read the script or decode the words, run the input as listening with pictures and do the deliberate study orally; if there is no time for input in class, set the reading or listening as homework and keep class time for steps 3, 6 and 7; if the learner's coverage is already high, skip step 2 and spend the deliberate share on the words their texts use rarely.

Fitting the design to a situation

The three facts that most change the decision:

  • How much of the learner's texts they already understand. Below about 95% coverage, inference from context fails and the intentional share must be larger; above it, input can carry more of the load. If unstated, ask: in a typical text the learner will read or hear next, how many words in a hundred do they not know?
  • How much meaning-focused input the learner gets, and where. A course with an hour of reading a week can rely on input more than one with ten minutes; input outside class (a phone, a library, work documents) counts. If unstated, ask: how many minutes a week does the learner read or listen to the language for meaning, in and out of class, and on what?
  • What the learner must do with the words, and when. Recognising a meaning, recalling it, and using a word in speech or writing at a delay need different shares: productive use needs the after-input routine and spaced use. If unstated, ask: what should the learner be able to do with these words, and how long after the course?
If the brief says… Then change… Basis
Learners are novice in the language (beginners, below about the first thousand words) Lead with deliberate study (about two thirds of vocabulary time); use graded or simplified texts with most unknown words glossed; check coverage on a 100-word sample before any text Correct word inference in reading requires knowing roughly 95-98% of surrounding words… [+M] (second-hand, one review); ratio untested
Learners are advanced, already at about 95% coverage of their texts Shift most time to extensive reading and listening with few glosses; keep deliberate study for low-frequency words in their field and for the after-input routine Incidental vocabulary exposure produces reliable learning, but learners pick up only a small share… [~S] (why some deliberate study stays even for strong readers); ratio untested
Learners are child first-language readers, not second-language learners Keep the same combination: read aloud and talk about texts, pre-teach two or three words, and reuse them in talk; expect taught-word gains Pairing Contextual Encounters… [+M] (a meta-analysis of vocabulary instruction, population not stated in the entry); most other claims here are second-language and university, so carried
Adults in work settings learning a job's terms in a second language Take the input from their real documents, manuals and conversations; pre-teach the terms each document needs; keep a team word list from what they meet; check use in the job task untested; carried from university second-language evidence
The goal is productive use (speaking, writing), not only reading Keep the after-input routine for every text and add one speaking or writing task per week that needs the kept words; keep the sentence the word was met in Adult ESL students who read a short story could use incidentally acquired words in sentences… [~M] (second-hand, adult ESL)
Learners have low literacy, or the language uses a script they cannot yet read Run the input as listening (stories, audio, video with pictures) and the deliberate study orally with pictures and actions; add reading once decoding allows L2 learners pick up vocabulary incidentally from meaning-focused input… [+M] (listening yielded similar proportions of words learned to reading; viewing lower)
Setting is online-self-paced, phone-only, or no teacher present Put graded readings or audio with tap-to-see glosses in the app; schedule spaced recall of words the learner taps or saves; after each reading ask for one typed or spoken use of a saved word Listening to vocabulary lessons while physically active… [+M] (ESL students, phone audio added to class teaching); Supplemental computer-based spaced repetition… [+M] (EFL cadets, alongside classroom teaching, not instead of it); in-app glossing carried from Marginal glosses, dictionary use, and repeated occurrence… [+M] (printed text)
Some learners are remote and some in the room Share the text and its glosses on screen or in a shared document; take after-input uses in the chat as well as aloud; set the extensive reading as individual work both groups do the same way untested proposal
One teacher with a large group, or scarce facilitator time (a volunteer, a short weekly check-in) Pre-teach to everyone at once; let input and spaced review run on readers, audio or an app; in the check-in look at one thing: two or three uses of kept words from the week's reading untested proposal; the review part rests on Supplemental computer-based spaced repetition… [+M]
A single session only, or a fixed short time budget (for example 30 minutes) 5 minutes pre-teaching 3 words, 15 minutes reading a short glossed text where they recur, 5 minutes keeping 3 words with their sentences, 5 minutes using them; give the words on a card for later review; do not claim retention untested proposal
Learners have low confidence or anxiety about not understanding Start with texts at 98% coverage or more and glossed; let learners mark unknown words privately; make the after-input routine written before spoken untested proposal; the 98% figure from Correct word inference in reading requires knowing roughly 95-98%… [+M]
The brief names its own routine (an extensive reading programme, a vocabulary log, a word wall, a set textbook glossary) Keep the routine; change only what the evidence says matters: check coverage of the texts, pre-teach a few words before each, gloss in the margin, and bring kept words back on a schedule Combining intentional and incidental vocabulary learning… [+M]; Incidental vocabulary gains show retention problems… [~M]
The brief gives its own word list or categories (a course list, a frequency band, an exam list, a job glossary) Teach from that list in its own categories; choose input texts in which its words recur, rather than replacing the list untested
Learners keep a private reading log or word journal The learner decides what goes in it and who sees it; check learning through a separate use task, not by reading the log untested proposal
The output is a performance or product for a real audience (a presentation, a client conversation, a published text) that the brief wants kept whole Feed the performance with reading and listening on its topic and pre-teach its key words; rehearse parts of the real task; let the performance run whole rather than becoming a vocabulary test untested proposal
Stakes are high (an exam with a set word list, a reading test in a fixed time) Raise the deliberate share for the listed words, with spaced recall until the exam; keep input at high coverage so reading speed grows; do not rely on input to supply listed words Intentional vocabulary learning yields higher, faster gains… [+M] (second-hand, one direct comparison)

Ten of the sixteen rows cite a claim for their main move or a part of it (novice, advanced, child, productive use, low literacy, self-paced, scarce time, low confidence, named routine, high stakes); the child, self-paced and advanced rows carry a claim to learners or settings it did not test and say so. The other six are untested proposals.

Observation, state and explanation

What can be observed: which words a learner marks as unknown in a text sample, their answers to comprehension questions, the glosses they open or the words they look up, the words they choose to keep after reading, their recall of a word's meaning at a review, and their use of the word in a sentence or task. "Knows the word", "has partial knowledge", "infers well from context" and "is not reading enough" are inferred states. A word recognised at the end of a session is not a word recalled two weeks later or used in writing.

Response Plausible explanations Observation that could change the interpretation
Understands the text but cannot give the meaning of new words in it afterwards Read for gist and skipped the words; context let them understand without resolving the word; glosses unused Ask for the meaning of three words in their sentence before and after pointing to the gloss; record gloss use.
Infers wrong meanings from context Too few known words around the target (low coverage); misleading context; carried a first-language meaning Count unknown words in the sentence and the paragraph; give a text at higher coverage with the same word and compare.
Recalls taught words in review but does not use them Recognition without production; no task that needed the word; the word met only in lists Give a speaking or writing task that needs the word, with and without a sentence frame.
Gains at the end of the unit, lost at two weeks Too few encounters; no spaced review; words met only in one passage Check how many times each word recurred in the input and in review between the tests.
Reads little outside class, and growth stalls Texts too hard (low coverage); texts not interesting to the learner; no time or access Sample coverage of the texts available, ask what they would read by choice, and record minutes read for a week.

These rows are proposals, untested with learners as diagnostics. A coverage check or a recall probe is itself an encounter with the word, so record each one.

Evidence and qualifications

Two claims test the combination and both are thin, and a third adds an input component to deliberate study in one small trial; the halves are better supported. Learners, outcomes and horizons differ across them (Chinese university ESL students, advanced Dutch university learners of French, adult ESL students, pooled second-language studies), so do not rank them by effect label.

What limits the model.

None of the claims varies the ratio of intentional to incidental time, the order of the two, or the coverage at which to shift; none follows learners across a course.

The evidence by question

Further evidence, not yet read against this model

Every claim the earlier page cited is used in the model above. Claims found while rewriting, read but not built into the model:

Objective and learner-valued goal

The designer's objective is usually a vocabulary large enough to read and listen to the course's texts at high coverage, plus known and usable target words at a stated horizon. Elicit separately what the learner wants: to read a particular kind of text (novels, news, manuals at work), to follow conversations, to pass a test with a set list, to talk about their own life. Agreement and divergence are both observations. A learner who wants to pass a listed-word exam will favour deliberate study and may not read; a learner who wants to read for pleasure may resist word lists. The input half is where the learner's own goal can most easily be served (choice of texts), and the evidence that interesting passages give longer-lasting growth is a reason to let them choose.

For example, an adult learner of English may want to read football news on a phone, while the course objective is the first two thousand words of a frequency list. Graded football reports with glosses, chosen by the learner, can serve both: the frequency-list words recur in them and are studied after reading. Record both the list words recalled and whether the learner reads the news unaided by the end, so success on one is not read as the other.

What would revise this model?

The expectation should weaken if primary studies comparing combined, intentional-only and input-only conditions with comparable learners, aligned outcomes and a delayed test find no advantage for the combination. If deliberate study of words, given with their contexts and reviewed on a schedule, is still forgotten as quickly as the focus-on-form words in the re-analysis above, the intentional share should shrink and its role change to glossing and noticing. If learners below 95% coverage turn out to learn as much from glossed authentic texts as from deliberate study of frequent words, the coverage switch-point should go. If a study varies the ratio or order of the two modes, its result replaces the default arc. Do not protect the model by calling every failure "not enough reading" or "not enough review" after the fact.

Words recognised at the end of a session, words recalled at a delay, words used in writing or speech, and growth in general reading are separate claims. The present evidence establishes no ratio, no order, no coverage switch-point tested as such, and no number of encounters per word.

Claims

Examples

- - Pre-teach, read, keep: 3–5 key words taught before a glossed graded reader chapter, then learners keep and use 3 words they met. - App-based extensive reading: graded stories with tap-to-see glosses, saved words returned on a spaced schedule.

Key Sources

  • Restrepo Ramos, F. D. (2015). Incidental vocabulary learning in second language acquisition: A literature review. PROFILE Issues in Teachers' Professional Development, 17(1), 157-166. http://dx.doi.org/10.15446/profile.v17n1.43957
Page metadata
Field Value
type principle
id combine-intentional-and-incidental-vocabulary-learning
title Combine intentional and incidental vocabulary learning rather than relying on incidental learning alone
description For second-language learners building general vocabulary, deliberate study of selected words combined with reading and listening for meaning (glossed, with recurring words) is expected to give more words and better retention than either mode alone, with the deliberate share largest while text coverage is below about 95%; two thin claims test the combination, none tests its order, ratio or dose.
status review
canonical True
generated by claude/unspecified on 2026-10-15
trust unverified — no one has confirmed this page's content
sources 1 — restrepo-ramos-2015

Frontmatter is the page's machine-readable record — see the Schema & Guide.