The Landscape Model of reading is a computational and psychological account of how concepts change in accessibility as a person moves through successive sentences. Developed by Paul van den Broek and colleagues, the model combines relatively automatic memory activation with the reader’s attempts to build coherent understanding and connects online processing to the memory representation that remains after the passage.
Did you know that a book mentioned near the beginning of a story can become relevant again many sentences later, even after the narrative changes scene? A repeated name, related action or causal clue can bring earlier information back into focus. Vocabulary works as a technology of connection when a reader links the right referent to the current sentence without inventing events that were never written.
This complete guide examines the 1996 original Landscape Model account, its computational implementation by Tzeng and colleagues in 2005, and later work on automatic retrieval, coherence seeking, prior knowledge and science-text misconceptions. It uses original cases for Primary and Secondary English, Mathematics, Science, History, General Paper, translation and professional records.
Editorial note: The cases and imagined activation patterns are original teaching illustrations, not newly measured participant data. A simulated concept-activation value is not a mind scan or diagnosis of any learner.
Contents: why concepts rise, fade and return while we read
Open 52 chapters and evidence checks
- 1. The Landscape Model studies changing concept activation
- 2. Vocabulary connects ideas across sentences
- 3. Van den Broek and colleagues developed the approach
- 4. A reading cycle is a modelling unit
- 5. Current text is the first activation source
- 6. Carry-over preserves recently available ideas
- 7. Reinstatement brings older ideas back
- 8. Background knowledge adds further connections
- 9. The four sources have different evidential status
- 10. Concept activation is dynamic
- 11. A concept can be active for different reasons
- 12. Repeated ideas can strengthen connections
- 13. Online comprehension differs from later memory
- 14. Tzeng and colleagues implemented the model computationally
- 15. A computational graph is not a mind scan
- 16. Memory-based activation can be rapid
- 17. Coherence-based processing can be goal-directed
- 18. The 2005 integration paper combined perspectives
- 19. Standards of coherence affect comprehension
- 20. A coherence search should not become invention
- 21. The same word can activate different senses
- 22. Semantic association is not identical to reference
- 23. A repeated person can revive an older scene
- 24. Pronouns need a defensible antecedent
- 25. Related nouns can refer to one object
- 26. A contradiction can cue old information
- 27. Inconsistent ideas can remain temporarily available
- 28. Science education provides a critical application
- 29. Causation can be inferred from related events
- 30. A goal can connect distant actions
- 31. Recent activation and lasting importance differ
- 32. An original library story illustrates the cycles
- 33. A five-column worksheet separates sources
- 34. A timeline prevents invalid reinstatement
- 35. A location shift should not erase the goal
- 36. Text layout can support coherent reading
- 37. Reader characteristics affect the pattern
- 38. The model does not prove individual intelligence
- 39. It differs from the Resonance Model
- 40. It differs from Construction–Integration
- 41. It differs from the Structure-Building Framework
- 42. It differs from event segmentation
- 43. It differs from event indexing
- 44. Primary English can practise reactivation
- 45. Secondary English can track an argument across paragraphs
- 46. Mathematics requires correct earlier quantities
- 47. History needs sources more than familiar associations
- 48. General Paper should test coherence against evidence
- 49. Translation needs stable reference across distance
- 50. Professional records require current-status checking
- 51. Simulations need explicit assumptions
- 52. The educational goal is a stable justified representation
Quick Test: four sources of useful reading information
| Information source | Where it comes from | Invented library example |
|---|---|---|
| Current text | Sentence currently being processed | Ben opened the reference book |
| Carry-over | Immediately preceding material | Ben remains the current character |
| Reinstatement | Older text made relevant again | Mira borrowed the book in the opening paragraph |
| Background knowledge | Prior knowledge of objects and situations | A reference book may support a school project |
These are conceptual categories, not invented activation measurements. A pupil should always distinguish what the author explicitly stated from general knowledge that could merely suggest a plausible event.
1. The Landscape Model studies changing concept activation
The Landscape Model explains how ideas can rise, fade and become active again as readers progress through a text. A character, place, object or cause might be central in one sentence, less prominent during another and then relevant when a later expression refers back to it.
The term landscape describes a changing pattern of activation rather than a physical picture of the brain. Its goal is to account for how comprehension unfolds during reading and how an integrated representation remains in memory afterwards. This makes it a theory of processing, not a pupil performance score.
2. Vocabulary connects ideas across sentences
A reader may know the meanings of borrowed, carried and returned yet misunderstand which person performed each action or which object was involved. Words provide forms and concepts, while discourse builds their relationships.
The model helps explain how a previously introduced item can be brought back into the current interpretation. A student who knows the vocabulary but loses the earlier referent has an integration problem rather than necessarily a dictionary problem. Teaching should identify that actual gap.
3. Van den Broek and colleagues developed the approach
Paul van den Broek, Kirsten Risden, Charles Fletcher and Richard Thurlow described a landscape view of reading in a 1996 chapter in Models of Understanding Text. They proposed examining how concept activation changes from one reading cycle to the next.
The title emphasised fluctuating activation and the construction of a stable memory representation. The research did not claim that all concepts remain equally accessible at every sentence. Later publications expanded the account with computational implementations and comparisons with actual reading and memory data.
4. A reading cycle is a modelling unit
A simulation may treat each sentence or proposition as one cycle. Incoming material at that step contributes to the activation of concepts already represented in the model and to the retrieval of other information.
Real readers do not process language in rigid computer ticks. The cycle is a theoretical simplification allowing researchers to track activation changes and make predictions. An educational diagram should not imply the actual brain follows the software clock with a perfectly uniform rhythm.
5. Current text is the first activation source
The sentence currently being read can activate the people, objects and actions that it explicitly describes. If a story says Ben opened a book, concepts representing Ben, opening and the book are relevant to that cycle.
These concepts are not necessarily identical to the individual printed words. The same object may later be described as the volume or it. Correct comprehension requires mapping the various expressions to the intended referent.
6. Carry-over preserves recently available ideas
Information from the immediately preceding reading segment may remain active as the next sentence arrives. This continuity helps a reader interpret pronouns and related actions without requiring complete repetition.
After ‘Mira borrowed a book’, the sentence ‘She carried it into the classroom’ can be understood because the character and object remain available. The reader must still verify the antecedents. The model concerns availability, not a guarantee that every pronoun is unambiguous.
7. Reinstatement brings older ideas back
A detail from an earlier passage may become relevant again after several intervening sentences. A new word or related action can reactivate that older information even when it is no longer the immediate focus.
For example, a book introduced in the first paragraph may matter again when a character finally begins a presentation. The student should identify which earlier item the later sentence retrieves. A useful reading strategy is to link the two textual mentions explicitly.
8. Background knowledge adds further connections
Readers possess knowledge about ordinary activities such as visiting libraries, preparing lessons and using books. The current sentence can activate some of those familiar ideas alongside information from the passage itself.
This background knowledge can support inference but can also mislead. A library visit does not prove someone borrowed a book, and an examination does not always mean the learner was nervous. Students should distinguish ordinary associations from facts the author actually supplied.
9. The four sources have different evidential status
The model distinguishes information in the current text, carry-over from recent processing, reinstated material from earlier reading and background knowledge. These sources may overlap in the concepts they activate.
Yet directly quoted text has a different status from a plausible inference based on general experience. A classroom worksheet should mark the source of each claim. Strong vocabulary skill includes not only noticing a connection but knowing whether the connection is stated or inferred.
10. Concept activation is dynamic
When the narrative moves away from a book toward a different subject, the book may lose immediate prominence. When the narrator mentions that same book again, related information can become accessible once more.
That fluctuation is not permanent forgetting or relearning of the word book. It concerns what information is useful for the ongoing discourse. A student who confuses these concepts may believe every drop in activation means the reader has lost vocabulary knowledge.
11. A concept can be active for different reasons
A book may be salient because it is named in the sentence, because a pronoun refers to it, because the story is about preparing an essay or because the reader has just recalled a similar experience.
The model asks how these sources combine over time. A single high activation value does not tell the reader which cause was responsible or whether the idea was correctly interpreted. Experimental designs and coding rules must separate relevant influences.
12. Repeated ideas can strengthen connections
Concepts that recur and are coactivated across reading cycles may become linked in the developing memory representation. A central goal can connect distant actions even when the surface wording changes.
The amount of repetition alone does not establish importance. An irrelevant word might occur often, while one pivotal decision appears only once. The model considers the pattern of relations rather than offering a simple rule that the most frequent noun is the main idea.
13. Online comprehension differs from later memory
Online measures address processing while the reader encounters the text. Offline measures can ask what the person recalls, recognises or understands after the passage ends.
The Landscape Model attempts to link these phases. A student might read each sentence quickly but later forget which actor was responsible for an action. Another may read slowly and remember the causal explanation accurately. Speed and understanding are distinct outcomes.
14. Tzeng and colleagues implemented the model computationally
Tzeng, van den Broek, Kendeou and Lee published a computational implementation of the Landscape Model in Behavior Research Methods in 2005. Their paper described an architecture and parameters for representing online reading cycles and later memory.
They compared simulations with behavioural observations across different reading conditions and types of text. This is an empirical modelling programme, not simply a diagram. The software remains a model whose predictions depend on the assumptions and inputs chosen.
15. A computational graph is not a mind scan
A graph can show the modelled activation of the book concept across sentences. Peaks and declines illustrate how the theoretical system responds to input and connections.
Those lines are not EEG or fMRI recordings unless a particular study explicitly collected and analysed such data. A teacher drawing an imaginary curve should label it as a conceptual illustration. The value lies in explaining a hypothesis, not inventing scientific numbers.
16. Memory-based activation can be rapid
One research tradition proposes that current words cue related earlier material automatically or with relatively little deliberate effort. A name or associated feature may help previous narrative information become available.
This does not mean the retrieval is always accurate or relevant. A concept can be activated because it shares a feature with the present sentence even when it does not explain the current event. Accessibility is one step, not proof of a valid inference.
17. Coherence-based processing can be goal-directed
A reader may actively search for an earlier cause, resolve a pronoun or reread a sentence when the passage lacks an obvious connection. Such processing is directed toward achieving a sufficiently coherent interpretation.
The difficulty and necessary depth depend on the task. Scanning a train timetable requires different coherence goals from interpreting a scientific explanation. A strong model should accommodate those differences without turning every pause into a sign of insufficient vocabulary.
18. The 2005 integration paper combined perspectives
Van den Broek, Rapp and Kendeou published Integrating Memory-Based and Constructionist Processes in Accounts of Reading Comprehension in 2005. They argued that relatively automatic memory activation and readers’ efforts to establish coherent meaning interact during reading.
The Landscape Model was presented as a framework for combining these influences. It does not imply that every automatically retrieved idea is useful or that every deliberate explanation is correct. Readers still need to test their claims against the text.
19. Standards of coherence affect comprehension
Readers differ in how much explanation they seek. A reader preparing for a literal comprehension quiz may track names and stated events, while one studying a cause-and-effect essay may seek a more detailed mechanism.
A standard for coherence describes the kind and strength of connections the reader aims to establish under the reading goal. The educational implication is to make the task clear. A learner cannot meet a high-stakes accuracy requirement merely by grasping the general topic.
20. A coherence search should not become invention
When a story contains a gap, readers can use knowledge to propose a bridge. Mira might leave the library and later arrive in a classroom; the reader can infer travel occurred.
But the passage does not license a specific bus route or travel time if none was supplied. Coherence can be achieved without filling every detail. Good learning distinguishes necessary inference from an attractive imaginary addition.
21. The same word can activate different senses
The word bank can refer to a financial institution or a river edge. A text about flooding favours the geographical sense, while a text about payments favours the financial one.
A model of discourse activation can represent the availability of related concepts, but a student must still select the meaning supported by the sentence. Vocabulary becomes a technology of contextual interpretation rather than permanently memorising one meaning for every form.
22. Semantic association is not identical to reference
Rain and umbrellas are associated because they often occur together. If a story mentions a storm, umbrella-related knowledge may become active even though no umbrella is present.
An accurate summary must not add an umbrella as a narrated object without evidence. The model helps explain how an association arises, not why it should be treated as fact. This distinction is crucial for reliable school answers.
23. A repeated person can revive an older scene
A character may disappear for several paragraphs while the narration follows someone else. When the first character returns, related earlier events can become relevant again.
The student should determine whether the same person is being described and which earlier facts remain true. A name recurrence alone cannot establish that every old circumstance remains unchanged. Situation-model updating and reinstatement can occur together.
24. Pronouns need a defensible antecedent
A pronoun such as she can cue a previous female character, but a passage containing several possible referents may remain ambiguous.
A reader cannot justify choosing an antecedent simply because its name is highly familiar. Grammar, discourse focus and context determine which interpretation is supported. A reference chart provides a practical school method for tracking the intended link.
25. Related nouns can refer to one object
An author might introduce a telescope and later refer to the instrument. If the story contains only that telescope, the second expression may reasonably reinstate the first object.
But a Science laboratory can contain multiple instruments, so the reference needs checking. Reinstatement is useful when it reconnects the correct entity. A vague similarity in category should not create a false object identity.
26. A contradiction can cue old information
An earlier sentence may describe a character as unable to swim, while a later sentence says they crossed a river without a vessel. The newer claim could make the earlier limitation accessible again.
A reader should examine whether the text supplies a new explanation, such as a rescue or a shallow ford, before announcing an error. Memory activation helps detect potential inconsistency but does not settle every story detail automatically.
27. Inconsistent ideas can remain temporarily available
A learner may have a prior misconception that a passage seeks to correct. Reading the new scientific explanation can activate the old idea because the same concepts or words are involved.
Such activation does not indicate that the misconception has been accepted. To establish learning, the teacher needs a new question testing whether the pupil now applies the correct mechanism. Familiarity and belief are not interchangeable.
28. Science education provides a critical application
Van den Broek’s 2010 review in Science discussed the importance of coherent mental representations that combine text information with relevant background knowledge. It also considered how misconceptions can interfere with learning from texts.
A correct term can be memorised without correcting an underlying conceptual error. Vocabulary teaching should include explanation, counterexamples and transfer. The Landscape Model offers a way to discuss why earlier knowledge matters, not a guarantee that simple exposure cures misconceptions.
29. Causation can be inferred from related events
A passage might describe a glass falling and breaking. A reader can use common knowledge of fragile objects to infer a causal relationship even when the writer does not explicitly say because.
Such inferences help make the discourse coherent but depend on the circumstances. A metal cup might not break in the same way. Students should distinguish the narrated sequence, possible mechanism and strength of the evidence.
30. A goal can connect distant actions
A character planning a school presentation may search for a source, take notes and later design slides. The common goal explains why events separated by several sentences belong together.
When a later sentence mentions the presentation, earlier actions may become relevant again. However, preparing slides does not prove the presentation was delivered. The model helps organise intentions and events while the text determines which outcomes actually occurred.
31. Recent activation and lasting importance differ
An unimportant object might be named in the current sentence and therefore be highly accessible. A central cause from an earlier chapter might be less immediate but still essential to understanding the whole story.
A reader summarising the passage should focus on explanatory importance rather than merely repeating the most recently mentioned nouns. This distinguishes immediate access from the stable representation that the Landscape Model also seeks to explain.
32. An original library story illustrates the cycles
Imagine: ‘Mira borrowed an astronomy book. She visited the classroom. Ben prepared a presentation. Mira returned to the library. The librarian asked about the astronomy book.’
The book is directly mentioned in the first cycle, less central during the intervening classroom events, and mentioned again at the end. A teacher can ask how the final noun phrase connects to the earlier object. This is a qualitative illustration, not a computed graph.
33. A five-column worksheet separates sources
Use columns headed current proposition, carry-over, reinstated earlier material, relevant background knowledge and checked interpretation. For each sentence, identify the ideas that could contribute to comprehension.
The final column matters most: it should contain what the text actually supports. The sheet need not assign invented activation percentages. Students learn more from justified relationships than from false numerical precision.
34. A timeline prevents invalid reinstatement
An earlier book may have belonged to Mira before being passed to Ben. Later mention of the book should reinstate its identity without restoring the outdated belief that Mira still holds it.
The learner must integrate retrieval with updated state. This illustrates why memory and situation-model updating are related but not identical. An old fact may need to remain historically accessible while its present-tense status has changed.
35. A location shift should not erase the goal
Mira may leave a library and go to a classroom while still preparing the same presentation. Space changes but the goal remains continuous.
A good reader can update the current location and retain the overall intention. The Landscape Model can represent the changing relevance of these concepts, while the event-indexing model names the dimensions. Distinguishing them leads to a stronger explanation.
36. Text layout can support coherent reading
Paragraph breaks, headings and carefully chosen transitions can signal which earlier concept the new material elaborates. Such cues may make useful relationships easier to recover.
A heading is not itself proof that the paragraph is accurate. The student should identify whether the claims, examples and qualifications genuinely belong under that heading. Clear organisation supports comprehension without replacing evidence.
37. Reader characteristics affect the pattern
Prior vocabulary knowledge, relevant subject expertise, working-memory constraints and reading goals can affect which concepts are activated and used during comprehension.
A pupil deeply familiar with a topic may retrieve different examples from another pupil. Neither should be judged by one hypothetical activation diagram. The correct educational assessment asks whether each can understand and apply the actual subject information accurately.
38. The model does not prove individual intelligence
A computer simulation can fit patterns of concept activation and memory retrieval across groups of experimental materials. That achievement does not turn the model into a clinically validated test of a student’s intelligence or learning condition.
Research findings should be reported with participants, tasks and uncertainty. Classroom exercises are useful when they improve comprehension, not when they claim to infer an invisible personal activation score from an ordinary worksheet.
39. It differs from the Resonance Model
Myers and O’Brien’s resonance account emphasises the rapid, largely memory-based reactivation of earlier related information when new discourse cues are encountered.
The Landscape Model includes a memory-based route but also represents coherence-seeking processes and the development of a stable memory structure over reading cycles. One focuses especially on access to earlier discourse; the other attempts to model a wider dynamic comprehension process.
40. It differs from Construction–Integration
Walter Kintsch’s Construction–Integration model distinguishes the broad construction of candidate meanings and propositions from their later integration into a coherent network.
The Landscape Model focuses on the changing activation of concepts across successive cycles and resulting memory connections. Both involve networks and coherence but are not the same algorithm. A research article should cite the proper architecture and its specific hypotheses.
41. It differs from the Structure-Building Framework
Morton Ann Gernsbacher’s Structure-Building Framework describes laying foundations, mapping compatible information and shifting to new substructures, with enhancement and suppression mechanisms.
The Landscape Model instead foregrounds activation patterns, multiple retrieval sources and memory over time. A story shift can be discussed using both frameworks, but the scientific explanation changes according to which model is being tested.
42. It differs from event segmentation
Event segmentation concerns how continuous activity is divided into meaningful episodes. A person might mark a boundary when a character changes locations or goals.
The Landscape Model asks which concepts become active at each point and how connected information contributes to final memory. A strong reader can mark the event boundary and also retrieve the relevant earlier cause. Those are complementary skills.
43. It differs from event indexing
Event indexing tracks continuity through time, space, protagonist, causality and intention. These dimensions help describe relationships among narrative events.
The Landscape Model concerns activation dynamics and memory connections, not merely the list of narrative dimensions. A student can use an event-index table to identify what changed while separately asking which earlier information the text brought back into focus.
44. Primary English can practise reactivation
A short story introduces a missing pencil, follows a playground episode and later returns to the pencil. Ask pupils where the object was last seen and which words signal its return.
The lesson targets reference and memory in an accessible way. It does not need a computer simulation or neuroscience language. Correct answers should cite the text and distinguish what is known from what the child merely expects.
45. Secondary English can track an argument across paragraphs
An essay may state its thesis, give a long example and later return to the original claim. Students who concentrate on the newest example may lose the governing proposition.
Ask them to write the original thesis in a margin note and connect the concluding paragraph back to it. Then test whether the evidence genuinely supports the claim. Retrieval alone does not make an argument valid.
46. Mathematics requires correct earlier quantities
A word problem can introduce twelve objects, describe a transfer and later refer to the original quantity. The student must retain the earlier value while updating the present amount.
A state table helps separate original and current quantities. The Landscape Model offers a conceptual account of why some earlier information may be retrieved, but the correct answer still requires accurate numerical relationships and calculation.
47. History needs sources more than familiar associations
A later paragraph may mention an official previously described during an earlier event. The reader should retrieve the correct person, date and institutional role rather than attach a familiar but wrong name.
Historical evidence determines who acted. The accessibility of a concept does not make it true. A good summary distinguishes primary documentation, later interpretation and any unsupported inference.
48. General Paper should test coherence against evidence
A fluent editorial may repeat familiar terms and make related ideas highly accessible, creating the appearance of a coherent argument.
Students need to test premises, causal warrants and counterexamples. A psychologically easy interpretation can still be factually or logically weak. The Landscape Model helps explain accessibility, while critical reasoning determines whether the conclusion is justified.
49. Translation needs stable reference across distance
A legal or technical document may introduce one organisation and refer back to it later through an abbreviation. The translator needs to reinstate the same entity without changing ownership or decision authority.
An entity ledger can help preserve consistency across paragraphs. The model explains why retrieval cues matter, but fidelity must be verified directly against the source. Smooth target-language prose is inadequate if it changes the referent.
50. Professional records require current-status checking
An incident report can introduce an unresolved fault, describe several investigations and later state a repair decision. The reader must retrieve the original fault and distinguish the plan from the completed repair.
A strong report uses identifiers, times and precise status words. Mental continuity does not confer authorisation. The current authoritative record, not a plausible narrative, controls whether the action may proceed.
51. Simulations need explicit assumptions
A computational implementation must define how text is segmented, which concepts are represented, how associations are weighted and how activation changes across cycles.
Changing those assumptions may alter the output. A graph that lacks model parameters or a comparison with behavioural evidence is merely illustrative. Students should learn to demand a clear method before accepting a numerical claim about comprehension.
52. The educational goal is a stable justified representation
After reading a passage, a learner should be able to identify the important people, objects, events and causal links and explain how the latest information relates to earlier text.
This is a stronger demonstration than identifying isolated vocabulary words. The Landscape Model suggests why multiple memory sources influence such understanding, but the final interpretation must remain anchored in the actual passage and the best available evidence.
Full worked case: a book reappears after several events
An original story begins: ‘Mira borrowed a book about planets.’ Several sentences describe her journey to a classroom and a conversation with Ben. The later sentence ‘Ben opened the reference and showed Mira the page about Jupiter’ can reactivate the earlier book, provided the story establishes that it is the same item.
Students should point to the earlier expression and justify the later reference. They must not invent an unreported route or additional copy of the book. This is the practical skill of reinstating a relevant concept while remaining faithful to the actual text.
Eight-stage text-connection workbook
- Identify the main proposition and referents in the current sentence.
- Record relevant information carried over from the immediately preceding sentence.
- Search earlier text for a concept that the current wording could reactivate.
- Mark any background knowledge used to make a bridging inference.
- Distinguish explicit reference, justified inference and loose association.
- Update locations, ownership and goals when the story changes state.
- Produce a faithful summary and verify every factual relationship against the passage.
- Write a fresh original story that brings back one earlier concept coherently.
Frequently asked questions
1. What is the Landscape Model?
A model of fluctuating concept activation during reading and its contribution to coherent later memory.
2. Who developed it?
Paul van den Broek and colleagues, beginning with their 1996 landscape account and later computational research.
3. What is a reading cycle?
A modelled processing step for a sentence or other specified text segment.
4. Which information sources are included?
Current text, carry-over, reinstated older text and relevant background knowledge.
5. What is memory-based activation?
Availability of related information triggered by current linguistic cues, sometimes without deliberate retrieval.
6. What is coherence-based processing?
Effort toward finding meaningful connections under the reader’s comprehension goals.
7. Is the model identical to event indexing?
No. Event indexing tracks narrative dimensions; the Landscape Model explains activation and memory dynamics.
8. Does a strongly activated thought have to be true?
No. Irrelevant associations or misconceptions can also become available.
9. Does the theory measure pupils’ brain activity?
No. A modelled concept activation pattern is not an individual neural recording.
10. What should students learn?
Reactivate the relevant earlier ideas, update the state and preserve evidence-supported meanings.
Original scholarship and connected vocabulary articles
- Van den Broek, Risden, Fletcher and Thurlow: A Landscape View of Reading — original research chapter and catalogue record.
- Tzeng, van den Broek, Kendeou and Lee (2005): The Computational Implementation of the Landscape Model — running example and simulated cycles.
- DOI for Tzeng and colleagues (2005) — stable scholarly identifier.
- Van den Broek, Rapp and Kendeou (2005): Integrating Memory-Based and Constructionist Processes — combined processing account.
- Rapp and van den Broek (2005): Dynamic Text Comprehension — fluctuations in activation and memory.
- Van den Broek (2010): Using Texts in Science Education — prior knowledge, misconceptions and coherence.
- Learning from Texts: Activation of Information from Previous Texts During Reading — later cross-text retrieval evidence.
- Vocabulary | What Is the Construction–Integration Model? — a distinct comprehension architecture.
- Vocabulary | What Is the Structure-Building Framework? — foundations, mapping and shifting.
- Vocabulary | What Is the Event-Indexing Model? — narrative event dimensions.
- Vocabulary Learning Hub — develop context-aware language knowledge.
The Landscape Model shows that comprehension develops dynamically: earlier ideas can fade and return as the reader uses new language, memory and coherence goals. Its educational value is the discipline to recover the right relationship, discard unsupported guesses and form a meaningful account of the whole passage.