Semantic priming is the change in processing a target word or concept produced by a preceding, meaningfully related stimulus. A familiar example is recognizing nurse more quickly after doctor than after an unrelated word such as window. In an experiment, doctor is the prime and nurse is the target. Researchers measure the difference between carefully matched related and unrelated conditions using response times, accuracy, or brain activity. Semantic priming provides one of psycholinguistics' most productive methods for investigating how lexical knowledge, meaning, context, and attention interact. Meyer and Schvaneveldt (1971); Mangat et al. (2026).
The experimental phenomenon is well established, while its mechanisms require precise interpretation. A related word can facilitate processing through learned associations, shared semantic features, spreading activation, prospective expectations, or postlexical response strategies. The same apparent benefit may arise at different cognitive stages depending on the experiment. A rigorous explanation therefore begins by separating the measured effect from a theory about what produced it. Hutchison (2003); Mangat et al. (2026).
What Is Semantic Priming in Psychology?
Semantic priming is operationally defined through how a prime changes performance on a subsequent meaning-related target. If related targets are recognized faster or more accurately than unrelated targets, researchers describe a facilitation effect. A priming effect can also take the form of altered eye movements or a different event-related brain potential. This definition identifies an observable comparison; it leaves open the question of which mental operations explain it.
Consider a purely illustrative calculation. If a group identifies a word following an unrelated prime in 620 milliseconds on average and the same target following a related prime in 590 milliseconds, the estimated facilitation is 30 milliseconds. These values are an example of arithmetic, not an empirical effect size. Interpretation still depends on whether prime words, target words, and response contingencies are comparable and whether the experimental design permits a valid causal inference about the manipulated relationship.
Priming also has a direction. Bird may bring robin to mind more often than robin brings bird to mind, and asymmetrical association can affect experiments that reverse word order. Strongly associated pairs are not identical to pairs with high conceptual similarity. A hammer and nail share an event, whereas a robin and sparrow share category membership. Conflating these relationships has been a major source of competing interpretations. Lucas (2000); Hutchison (2003).
Semantic similarity, association, and thematic relations
Semantic similarity refers to overlapping conceptual attributes or position within a shared category. Association refers to the probability that encountering one item evokes another, frequently estimated from free-association norms. Thematic relations connect objects or actions that participate in the same event. Bread and butter are thematically related and often associated; dog and cat are taxonomic neighbors and commonly associated; key and lock are functionally related. Real stimuli regularly possess several kinds of relatedness simultaneously.
Research can separate these dimensions by collecting feature norms, association probabilities, corpus-based co-occurrence measures, and judgments of thematic or categorical similarity. Such measurements are imperfect, but they permit questions that intuitive example pairs cannot resolve. A pair labeled semantically related is an experimental claim that requires independent justification. McRae et al. (2005); Hutchison (2003).
Semantic priming versus repetition priming
Repetition priming occurs when the same stimulus appears again, as in nurse followed by nurse. Repetition may facilitate perceptual encoding, orthographic identification, lexical access, or episodic retrieval. Semantic priming involves a meaningful relationship between distinguishable stimuli, as in doctor followed by nurse. Form priming, meanwhile, depends on shared spelling or phonological properties even when meaning differs. These experimental categories can interact, but each makes a different theoretical claim. Forster and Davis (1984).
How Semantic Priming Became a Central Research Paradigm
Meyer and Schvaneveldt's 1971 work demonstrated that decisions involving associated word pairs were facilitated compared with decisions involving unassociated pairs. The findings stimulated theoretical work on how retrieval operations for separate words become dependent on one another. Collins and Loftus subsequently proposed spreading activation as an influential framework for semantic processing, and Neely's experiments distinguished relatively rapid activation from slower attention-dependent expectations. Meyer and Schvaneveldt (1971); Collins and Loftus (1975); Neely (1977).
Subsequent research complicated the early picture. Studies varied the proportion of related trials, stimulus-onset asynchrony, decision requirements, masking, and the semantic relationship itself. Meta-analyses and critical reviews showed that association and feature overlap can both contribute, but that not every nominally related category pair produces the same facilitation. A comprehensive 2026 review treats priming as the outcome of interacting lexical, relational, modality, and task factors. de Groot (1984); Lucas (2000); Mangat et al. (2026).
The Structure of a Semantic Priming Experiment
A standard trial presents a fixation marker, a prime, a target, and a request for a response. In a visual lexical decision experiment the participant judges whether the target letter string is a real word. Related and unrelated trials are counterbalanced across people and lexical items. Often a participant sees each target only once, preventing simple target repetition from dominating the result. Accuracy and latency are recorded, with appropriate treatment of errors and extreme response times.
The prime must be distinguished from the target, the relation between them must be defined, and the experimental conditions must differ in a way that isolates the intended causal factor. Counterbalancing prevents a particular set of inherently easy words from appearing only in the related condition. Materials matched on target frequency, length, spelling neighborhood, familiarity, and concreteness strengthen the interpretation. Norming word pairs in the relevant linguistic population helps ensure that the proposed relationships are actually perceived. McRae et al. (2005); Mangat et al. (2026).
Lexical decision tasks
Lexical decision requires a word-or-nonword judgment about the target. It is widely used because responses are easy to collect and often sensitive to semantic relationships. However, the response is a decision, not a direct readout of lexical activation. When related words disproportionately receive a positive response and nonwords are usually unrelated, participants may exploit relatedness as evidence that the target is a word. Careful designs balance response contingencies and investigate whether the effect survives changes in task. Meyer and Schvaneveldt (1971); Kinoshita and Norris (2012).
Word naming and pronunciation
Naming experiments ask participants to pronounce the target quickly. This shifts the demands toward visual recognition, phonological retrieval, and speech initiation. Priming effects can be smaller than lexical-decision effects, and that discrepancy helps researchers separate effects associated with a word/nonword judgment from more general lexical-processing facilitation. Lucas's meta-analysis identified a relatively smaller mean effect in naming tasks among the paradigms considered. Lucas (2000).
Categorization and semantic verification
In semantic categorization the participant decides whether a target names an animal, an object, a living thing, or another specified class. Because the required response explicitly involves meaning, such tasks provide strong opportunities for semantic relationships to influence performance. They also introduce response preparation and category-specific strategies. A prime may bias the decision criterion even when it has not made the target's representation intrinsically easier to access. Mangat et al. (2026).
Masked priming
Masked priming briefly presents a prime between visual masks, limiting its reportable visibility. Researchers then test whether the target is affected. Meta-analytic evidence indicates that masked stimuli can influence subsequent performance, while visibility measures and task-specific mechanisms remain essential for interpretation. Masked effects do not establish comprehensive unconscious understanding of a prime's meaning; orthographic processing, partial awareness, response processes, and the particular mask sequence must all be considered. Van den Bussche et al. (2009); Kinoshita and Norris (2012).
Spreading Activation: The Classical Explanation
Spreading activation proposes that identifying one word activates a representation and increases activation of connected representations. When the target belongs to the activated neighborhood, less additional processing may be needed to recognize it. Connection strength, distance, and activation dynamics offer an intuitive explanation for graded semantic facilitation. Collins and Loftus developed an influential version of this account in their theory of semantic processing. Collins and Loftus (1975).
The theory is explanatory and testable, but its familiar network diagrams should not be treated as photographs of neurons or storage locations. A node is a theoretical representation, and an edge encodes a proposed relationship. Different kinds of conceptual knowledge may be represented by distributed patterns rather than discrete boxes. Similar observed priming can arise from several architectures, so evidence for facilitation is not by itself conclusive proof of spreading activation.
The model remains valuable because it predicts how relatedness, indirect connections, time, and competing activation might influence performance. It becomes scientifically informative when these predictions are compared with those of feature-based, retrieval-based, or decision-based alternatives. McNamara (2005); Mangat et al. (2026).
Distributed Semantic Features and Representations
Feature-based theories propose that related concepts overlap in semantic attributes. Dog and wolf may share biological, perceptual, behavioral, and functional properties. Processing one concept can make overlapping representational information more available for another. Distributed models represent meaning as patterns across many dimensions or units; activation need not move through a chain of separate lexical nodes.
The empirical challenge is identifying which features count and how strongly. Free-association strength, category membership, and feature overlap can covary. McRae and colleagues' semantic feature norms exemplify the kind of independent material characterization needed to test feature-sensitive predictions. Hutchison's review concluded that evidence supports contributions of both association strength and feature overlap under specific conditions. McRae et al. (2005); Hutchison (2003).
Compound-Cue and Decision-Level Accounts
Other explanations locate priming partly in the operation performed when the target appears. A prime and target may jointly form a more familiar retrieval cue, allowing a more efficient assessment of the pair. The participant may also judge the coherence of the pair after identifying the target. When a task rewards semantically coherent word pairs, this postlexical matching can accelerate a positive lexical decision.
Compound-cue and decision accounts emphasize that earlier information may help the response without first activating a specific target representation. Distinguishing this possibility from preactivation requires task comparisons, relation-proportion manipulations, neutral baselines, and processing measures independent of the final response. McNamara's synthesis discusses spreading activation, compound-cue, and distributed-network families together rather than treating any one as the default interpretation. McNamara (2005); de Groot (1984).
Automatic Processes and Strategic Expectations
Automatic semantic influence is usually investigated through relatively fast, low-intention, or attention-resistant effects. Controlled influence includes goal-directed expectancy: when bird is presented, a participant may predict particular bird names or animal categories. When this strategy helps on many trials, related targets can benefit and unexpected targets can incur costs. The experiment must distinguish these explanations through systematic manipulation. Neely (1977).
A rapid response alone does not certify an automatic mechanism. The prime might be visible and participants might anticipate a likely response, while automatic and strategic components can operate together. Conversely, information that changes an electrophysiological response without an explicit judgment can constrain decision-only accounts. An integrated interpretation specifies what was visible, expected, attended, and required by the task. Mangat et al. (2026).
Why relatedness proportion matters
Relatedness proportion is the proportion of relevant trials on which prime and target are related. If a large proportion is related, the prime becomes a useful forecasting cue. If related trials are uncommon, strategic prediction is less reliably rewarded. Experimental findings demonstrate that relatedness proportion and the prime–target temporal interval can interact. The resulting changes may reveal strategic adaptation rather than a fixed property of the lexical network. de Groot (1984).
Stimulus-onset asynchrony and processing time
Stimulus-onset asynchrony, or SOA, measures time from prime onset to target onset. The interstimulus interval instead measures the gap after prime offset. Shorter SOAs reduce opportunities for some elaborate anticipatory strategies, while longer SOAs allow additional operations. Neither timing variable uniquely identifies a mechanism. Prime duration, visibility, target modality, and participant expectations must be reported alongside it. Neely (1977).
How Strong Is the Evidence for Pure Semantic Priming?
The distinction between semantic and associative relationships has generated a productive dispute. Shelton and Martin found conditions in which automatic priming was evident for associated pairs but not for semantically related nonassociated pairs. Lucas's 2000 meta-analysis, examining 26 studies, reported evidence for semantic priming without direct association. Hutchison's 2003 microanalytic review challenged some interpretations of category-coordinate data and argued that both association and feature overlap influence the effect. Shelton and Martin (1992); Lucas (2000); Hutchison (2003).
These reviews do not establish that every related pair is processed automatically or that one experimental relation always dominates. Different item norms, tasks, time windows, and analytic decisions can change the pattern. The appropriate conclusion is that semantic relations can contribute to priming beyond direct association in suitable conditions, while the contribution of each relation type remains sensitive to the design and operational definition.
Category, Thematic, Functional, and Indirect Priming
Taxonomic relations involve shared category membership, as in robin–sparrow. Thematic relations connect participants in an event, as in key–lock. Functional relations concern uses and actions, such as scissors–cutting. Antonyms may share a dimension while conveying opposing values, and synonyms overlap semantically without being fully interchangeable. Indirect or mediated priming concerns a prime and target joined through an intermediate association.
Each relation can suggest a different pathway. A thematic pair may be related through learned event knowledge, while a category pair may share distributed semantic features. An indirect relation may support spreading-activation accounts, but a single positive result cannot specify which intermediate representation was involved. In many natural word pairs, association and conceptual relatedness coexist, requiring independent norms and careful claims. Hutchison (2003); Mangat et al. (2026).
The Role of Prime Frequency and Semantic Richness
Prime characteristics help determine what information is available when the target appears. Word frequency, familiarity, ambiguity, length, concreteness, imageability, and number of semantic features may change the depth or speed of prime processing. A word with many associates might activate a broad set of representations; an ambiguous word might bring multiple competing senses into play. The direction of an effect cannot be generalized without considering the target and task as well. Mangat et al. (2026).
Semantic richness is not one perfectly interchangeable variable. Researchers variously operationalize it through attribute norms, numbers of associates, senses, or contextual diversity. A claim that semantically rich words always prime more strongly would require direct evidence across these measurements. Better experiments model prime-level factors and relational factors together.
The N400: Electrophysiological Evidence About Meaning
Event-related potentials make it possible to measure semantic context effects without relying solely on the participant's final button press. The N400 is a negative-going component sensitive to meaning-related processing. Targets preceded by semantically related material frequently produce a smaller N400 response than comparable unrelated targets. The response is also affected by wider sentence and nonlinguistic context, showing that meaning-sensitive processing extends beyond isolated word pairs. Kutas and Federmeier (2011).
The N400 commonly develops within several hundred milliseconds after a stimulus appears, with a broad negative deflection often described around 400 milliseconds. Its precise measurement depends on electrode selection, time window, materials, and preprocessing. Crucially, an N400 difference is a neural measurement associated with meaning, not a direct index of a single mental operation or an anatomical location.
Does the N400 measure prediction or semantic integration?
Competing frameworks interpret N400 modulation in terms of semantic access, facilitated retrieval, or context-dependent integration, with predictions potentially contributing to those processes. A reduced N400 to a related target constrains a theory by demonstrating an early meaning-sensitive difference, but does not alone prove that the target was specifically predicted in advance. Mechanism claims require further designs that distinguish lexical association, contextual constraint, and response processes. Kutas and Federmeier (2011).
Semantic Priming in Speech, Pictures, and Sign
Priming research is broader than visual words. Spoken primes unfold over time and can alter recognition of subsequently heard or seen targets. Pictures can activate conceptual knowledge relevant to words, while signed language supplies a fully linguistic visual-manual route to meaning. The N400 is elicited by meaningful material across multiple sensory and representational formats. This cross-format sensitivity supports the importance of semantic processes rather than one unique orthographic mechanism. Kutas and Federmeier (2011).
The modalities nevertheless differ in perception and timing. Spoken words reveal phonetic information incrementally; a printed word may be presented all at once; a picture has no inherent spelling. The existence of a priming effect in several modalities does not establish that every early processing stage is shared. Experiments must distinguish cross-modal conceptual influence from perceptual overlap and task-specific expectations.
Semantic Priming in Bilingualism
Cross-language priming asks whether a word in one language changes processing of a word in another. Translation-equivalent priming can involve items such as dog and chien, which refer to the same concept across English and French. Cross-language semantic priming involves conceptually related but nonidentical referents, such as an animal word in one language followed by a related animal in another. Both offer tools for studying interaction across learned lexical systems.
A meta-analysis of 64 masked translation-priming experiments across 24 studies found evidence of noncognate translation priming in both first-to-second and second-to-first language directions; mean effects differed by direction in the analyzed samples. Such findings support cross-language interactions but must be understood in relation to proficiency, language dominance, presentation intervals, and the linguistic properties of the items. Wen and van Heuven (2017).
Studies with bilingual toddlers have reported translation and cross-language semantic priming, supporting the view that meaningful connections across language systems emerge early. These findings do not license a claim that all bilingual individuals have identical language organization or a generalized executive-function benefit. The evidence concerns language processing under specific conditions. Floccia et al. (2020).
Development and Individual Differences
A semantic relation is learned through repeated encounters with words, events, categories, and discourse. As vocabulary develops, the connections available to support priming can change. A child's performance may additionally reflect attention, reading experience, task comprehension, and general response speed. Direct application of adult lexical-decision norms to toddlers or early readers would therefore be methodologically unsound.
Adults also vary in vocabulary, linguistic history, familiarity with particular concepts, and cultural associations. Word pairs considered obvious by one participant population may be unfamiliar or differently associated in another. Researchers should obtain norms that represent the people being tested, analyze item variability, and refrain from treating the absence of a single priming effect as evidence that a person lacks semantic knowledge. Mangat et al. (2026).
Semantic Priming and Semantic Memory
Semantic priming reveals how previously activated meaning changes immediate processing. It therefore informs theories about the accessibility and organization of conceptual knowledge. The broader domain of semantic memory studies knowledge of facts, words, objects, properties, and concepts over the long term. A millisecond-scale target benefit and a long-term semantic representation are related scientific questions, but they are distinct outcomes requiring distinct evidence. McNamara (2005).
A neighboring phenomenon, the tip-of-the-tongue state, involves temporary inability to retrieve a familiar word despite a feeling of knowing it. Priming research can illuminate lexical accessibility, while tip-of-the-tongue research also studies retrieval failure and metacognitive experience. Neither effect can be reduced entirely to the other.
Implications for Sentence Comprehension
Natural speech and reading involve syntax, discourse history, context, and speaker expectations. An isolated prime–target pair deliberately strips away much of this structure to isolate a measurable influence. Semantic priming helps explain how prior meaning changes the availability of related information, but it is not a complete model of sentence comprehension or predictive processing.
An observed benefit may arise from preactivation before the target, easier semantic access after the target, or later integration of the target with prior context. Experimental contrasts between association strength, predictability, task relevance, and temporal measures are needed to distinguish these possibilities. A semantic priming effect does not establish that language comprehension is always driven by explicit prediction. Kutas and Federmeier (2011).
Clinical and Educational Boundaries
Semantic priming paradigms have been used to study lexical-semantic processing in people with aphasia and other language-related difficulties. The interpretation depends on the specific clinical group, symptoms, task, and comparison sample. Priming is an experimental effect rather than a diagnostic label, and a difference in priming performance cannot by itself determine a clinical diagnosis, type of impairment, or treatment choice. A systematic review of priming in aphasia has emphasized methodological variability and limitations on generalizing to assessment or intervention. Semantic priming in aphasia: a primer and a systematic review (2025).
Priming also should not be equated with lasting education. Briefly activating a related word can influence immediate processing without establishing durable vocabulary acquisition, reading improvement, or long-term retention. Instructional benefit requires trials that measure learning, delayed outcomes, and transfer beyond trained items.
Dyslexia and phonological processing require specialized scientific and clinical attention. Semantic-priming evidence does not substitute for research on reading disorders or for a validated assessment. The independently published Phonological Awareness and Dyslexia article examines condition-specific evidence.
Computational Models and the Artificial Era
Computational models may encode semantic structure through associations, vectors, distributed features, or context-dependent probability estimates. Researchers can test whether the model's predictions explain human priming magnitudes and their sensitivity to task, stimulus timing, and relationship type. Such comparisons are useful precisely when they generate falsifiable distinctions between competing hypotheses. McNamara (2005); Mangat et al. (2026).
Language models may reproduce semantic associations in generated text, but similar lexical behavior does not independently establish a human-like mechanism of understanding or subjective experience. A scientific comparison must specify the task, measurements, and limits of inference. The Artificial Era places human and artificial language systems within a wider historical field of study; the empirical phenomenon of semantic priming remains defined by the observations and models that actually support its analysis.
Methodological Limitations and Research Standards
A common confound is treating association, feature overlap, and thematic co-occurrence as interchangeable. Another is making the probability of a correct lexical decision depend on whether the pair is related. A third is interpreting masked effects without assessing prime visibility and the influence of orthographic structure. A fourth is generalizing an effect obtained from a small pool of English word pairs to all languages and populations. Hutchison (2003); Kinoshita and Norris (2012); Mangat et al. (2026).
Better designs use independently collected norms, balanced target sets, appropriate neutral and unrelated baselines, adequate participant and item sampling, transparent exclusion rules, and analyses that generalize beyond the particular word pairs selected. Researchers should report effect sizes and uncertainty, examine both speed and accuracy, and replicate theoretically diagnostic contrasts. Such practices improve the meaning of the estimate as well as its statistical reliability.
Evidence Status: Findings, Mechanisms, and Open Questions
Established findings: related semantic or associative context can change target processing under suitable experimental conditions. Behavioral studies, meta-analytic syntheses, and N400 research provide convergent evidence that access to meaning is context-sensitive. Timing, relationship structure, and task design influence the observed result. Meyer and Schvaneveldt (1971); Kutas and Federmeier (2011).
Supported but condition-dependent findings: semantic relations can facilitate targets even when direct association is minimized in some designs; translation equivalents can prime across languages; and masked primes can influence processing under specific visibility and task conditions. Effect sizes and interpretations vary. Lucas (2000); Wen and van Heuven (2017); Van den Bussche et al. (2009).
Contested mechanistic questions: the relative contributions of spreading activation, distributed feature overlap, compound-cue retrieval, expectancy, and postlexical matching; the degree of automaticity in each task; and the significance of N400 differences for competing theories of semantic access and prediction. Limited evidence does not support diagnosing disorders, guaranteeing educational improvement, or drawing conclusions about artificial subjective experience from priming-like behavior. Hutchison (2003); Mangat et al. (2026).
Frequently Asked Questions
What is the simplest semantic priming example?
Recognizing nurse after doctor may be faster than recognizing nurse after an unrelated prime. The effect must be evaluated in matched experimental conditions because familiar example pairs may differ in association strength and several other properties.
Is semantic priming the same as classical conditioning?
Semantic priming concerns the effect of meaningfully related prior information on target processing. Classical conditioning studies learning through relationships between events and their consequences or predictive significance. Semantic priming can draw on learned associations without being reducible to a conditioning paradigm.
Is semantic priming always unconscious?
No. Primes may be visible or masked. Some effects can arise with limited awareness, while others involve expectations and deliberate strategies. Prime visibility must be measured rather than assumed. Van den Bussche et al. (2009).
Does semantic priming prove spreading activation?
No single effect establishes a particular architecture. Spreading activation is one theory, and distributed semantic features, retrieval cues, and task-specific decisions can make overlapping predictions. Hutchison (2003).
How does the N400 relate to semantic priming?
The N400 is an electrophysiological response sensitive to the ease and context of meaning-related processing. Related targets often elicit a reduced N400 relative to unrelated targets. The effect does not uniquely specify whether meaning was predicted, accessed, or integrated. Kutas and Federmeier (2011).
Can semantic priming be used to diagnose aphasia or dyslexia?
Semantic priming is a research method and not a standalone clinical diagnostic instrument. Clinical questions require condition-specific, validated assessments and appropriately interpreted evidence.
Does semantic priming demonstrate long-term learning?
A shorter response latency measures immediate processing. Durable learning requires independent evidence of retention and transfer over time.
Scientific Conclusion
Semantic priming shows that meaning is processed in relation to what has just been encountered. Related primes can facilitate subsequent word recognition, influence semantic judgments, and modulate electrophysiological responses. These effects link lexical access, conceptual knowledge, learning, attention, and comprehension in a single powerful experimental paradigm.
A mature scientific explanation treats the effect as a measurement that requires a model. The relationships between words, the nature and duration of prime exposure, the cognitive demands of the task, and the properties of the target jointly shape what can be inferred. Evidence supports semantic context sensitivity and several complementary processes, while decisive claims about representational architecture and automaticity demand carefully discriminating experiments. Mangat et al. (2026).
Related Articles
Cluster Language, Thought and Psycholinguistics — Scientific foundations of language processing and conceptual understanding.
Semantic Memory: Facts, Meaning, and Knowledge — The long-term organization and retrieval of conceptual knowledge.
Tip-of-the-Tongue Phenomenon: Why Words Feel Just Out of Reach — Lexical retrieval difficulty and metacognitive experience.
Phonological Awareness and Dyslexia: What Research Shows — The specialized evidence concerning reading-related phonological skills.
Mental Lexicon: How Words Are Stored and Retrieved — the organization and retrieval of lexical knowledge.
References
Anjum, J., Niharika, M. K., Scoggins, K., and Apple, B. (2025). Semantic priming in aphasia: A primer and a systematic review. Aphasiology. DOI.
Collins, A. M., and Loftus, E. F. (1975). A spreading-activation theory of semantic processing. Psychological Review, 82(6), 407–428. DOI.
de Groot, A. M. B. (1984). Primed lexical decision: Combined effects of the proportion of related prime-target pairs and the stimulus-onset asynchrony of prime and target. Quarterly Journal of Experimental Psychology Section A, 36(2), 253–280. DOI.
Floccia, C., Delle Luche, C., Lepadatu, I., Chow, J., Ratnage, P., and Plunkett, K. (2020). Translation equivalent and cross-language semantic priming in bilingual toddlers. Journal of Memory and Language, 112, 104086. DOI.
Forster, K. I., and Davis, C. W. (1984). Repetition priming and frequency attenuation in lexical access. Journal of Experimental Psychology: Learning, Memory, and Cognition, 10(4), 680–698. DOI.
Hutchison, K. A. (2003). Is semantic priming due to association strength or feature overlap? A microanalytic review. Psychonomic Bulletin and Review, 10(4), 785–813. DOI.
Kinoshita, S., and Norris, D. (2012). Task-dependent masked priming effects in visual word recognition. Frontiers in Psychology, 3, 178. DOI.
Kutas, M., and Federmeier, K. D. (2011). Thirty years and counting: Finding meaning in the N400 component of the event-related brain potential (ERP). Annual Review of Psychology, 62, 621–647. DOI.
Lucas, M. (2000). Semantic priming without association: A meta-analytic review. Psychonomic Bulletin and Review, 7(4), 618–630. DOI.
Mangat, A., Taikh, A., Gagne, C., and Spalding, T. L. (2026). The influence of prime characteristics in semantic priming. Frontiers in Psychology, 17, 1680365. DOI.
McNamara, T. P. (2005). Semantic priming: Perspectives from memory and word recognition. Psychology Press. Publisher record.
McRae, K., Cree, G. S., Seidenberg, M. S., and McNorgan, C. (2005). Semantic feature production norms for a large set of living and nonliving things. Behavior Research Methods, 37(4), 547–559. DOI.
Meyer, D. E., and Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: Evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90(2), 227–234. DOI.
Neely, J. H. (1977). Semantic priming and retrieval from lexical memory: Roles of inhibitionless spreading activation and limited-capacity attention. Journal of Experimental Psychology: General, 106(3), 226–254. DOI.
Shelton, J. R., and Martin, R. C. (1992). How semantic is automatic semantic priming? Journal of Experimental Psychology: Learning, Memory, and Cognition, 18(6), 1191–1210. DOI.
Van den Bussche, E., Van den Noortgate, W., and Reynvoet, B. (2009). Mechanisms of masked priming: A meta-analysis. Psychological Bulletin, 135(3), 452–477. DOI.
Wen, Y., and van Heuven, W. J. B. (2017). Non-cognate translation priming in masked priming lexical decision experiments: A meta-analysis. Psychonomic Bulletin and Review, 24, 879–886. DOI.