RESEARCH DOSSIER
Culture and Language in Perceptual Judgments
Abstract. Perceptual judgments depend on sensory information, observer learning, task goals, response criteria, and culturally available linguistic categories. This fundamental research monograph examines their relationships through the history of linguistic relativity and comparative psychophysics, landmark experiments on color classification, bilingual language activation, hemispheric asymmetry, contextual scene viewing, event representation, spatial and temporal reference, and olfactory naming and evaluation. The evidence is organized as a critical narrative synthesis of original experiments, contemporary conceptual replications, cross-population research, and methodological critiques. The research supports measurable, domain-specific influences of language and culture on categorization, attention, memory, and some discrimination tasks while identifying disputes over whether the effect arises within sensory encoding, decisional access, or later reports. Ecological regularities, task procedures, universal sensory constraints, and within-population variation are integral to interpretation. The synthesis specifies experimental designs capable of isolating these mechanisms and outlines implications for cross-cultural measurement and the Artificial Era. No original experiment, formal systematic-review protocol, or meta-analysis is claimed.
Keywords: perceptual judgments; cross-cultural perception; language and perception; linguistic relativity; color categorization; psychophysics; categorical perception; bilingualism; verbal interference; visual attention; scene perception; motion events; spatial frames of reference; time representation; odor language; olfactory perception; response bias; measurement invariance; cognitive penetration; Artificial Era
Introduction: Perception as Culturally Situated Judgment
Perceptual judgment is an observer's response to sensory evidence under a task, a social setting, and an available repertoire of concepts and words. A participant who selects the bluer of two patches, recognizes a person in a scene, describes an object's trajectory, or rates a smell as pleasant must map physical stimulation into a decision. That mapping can depend on retinal or receptor signals, attention, memory, prior exposure, categories, and communicative expectations. The research object of this dossier is the contribution of culture and language to that mapping, including the circumstances in which such contributions are measurable and the conditions in which shared sensory constraints dominate. Treating judgments as multidetermined makes comparisons of color, space, scenes, motion, time, and smell empirically coherent without treating all their mechanisms as identical Witzel & Gegenfurtner (2018) Majid et al. (2018) Firestone & Scholl (2016).
The central questions concern causal location and scope. Do linguistic categories modify sensory discrimination, encourage differential allocation of attention, facilitate memory, change decision thresholds, or help people communicate perceptual distinctions? How do everyday ecological encounters and culturally acquired expertise interact with these effects? Cross-language differences in verbal responses alone establish differences in naming; carefully designed nonverbal tasks are needed to investigate other processes. Cross-population differences in accuracy or reaction time require comparable physical stimulation and response demands before an explanation by linguistic or cultural organization is defensible Kay & Kempton (1984) Papafragou et al. (2008) Henrich et al. (2010).
This monograph synthesizes experiments and critical scholarship rather than reporting new data. Original studies are connected across domains by their actual dependent measures: forced-choice selection, similarity ratings, delayed recognition, visual search speed, fixations, electrophysiological response, odor rankings, and verbal description. Published systematic syntheses and theoretical papers clarify interpretation but are assessed separately from observations. The organizing argument is that the strength of evidence for a cultural influence grows when the investigator identifies the mechanism, eliminates plausible response artifacts, and demonstrates transport across well-defined conditions. An influential finding may illuminate one decision paradigm while remaining unsettled as an account of sensory experience itself Cibelli et al. (2016) Witzel & Gegenfurtner (2011) Sorokowski et al. (2024).
Defining the Variables: Stimulus, Experience, Category, and Report
Physical stimulation is an experimentally presented pattern of light, sound, chemical compounds, motion, or spatial relations. Sensory encoding is the physiological and computational processing that preserves or transforms information about that stimulation. A perceptual estimate is an observer's inferred property of the environment; a category is a learned partition of possible estimates or events; a report is an observable choice, utterance, or action. These concepts describe different stages and may overlap in time. For instance, a color discrimination judgment can reflect a chromatic difference detected before any name is retrieved, while an explicit similarity judgment can draw on a learned categorical boundary. Identical verbal reports therefore do not imply identical encoding, and distinct verbal reports need not imply different sensations Winawer et al. (2007) Witzel & Gegenfurtner (2018) Cibelli et al. (2016).
Culture denotes a historically transmitted complex of practices, ecological arrangements, conventions, institutions, and shared learning experiences; it is measured through more specific exposures and practices rather than inferred merely from a passport. Language may operate as a set of lexical contrasts, grammatical patterns, habitual descriptions, online verbal strategies, or interpersonal signals. Bilingual speakers demonstrate that these dimensions can change within an individual, depending on language activation, proficiency, and immersion. Researchers must distinguish habitual category learning from immediate recruitment of a label during a trial, and both from education, visual environment, literacy, expertise, socioeconomic context, and experimenter interaction Athanasopoulos (2009) Sinkeviciute et al. (2024) Henrich et al. (2010).
A useful causal representation is stimulation plus observer history plus current task producing internal estimates, response selection, and observable behavior. Each arrow is an empirical hypothesis: labeling may influence a remembered estimate, response rule, or early prioritization; shared physiology may constrain every stage. This representation avoids prematurely choosing between a fully impermeable sensory module and unrestricted cognitive penetration. One can document a causal effect of words on judgment without establishing that photoreceptor processing or phenomenal color appearance changed. Conversely, stable low-level sensitivity does not exclude genuine influences on attention or the information made available for categorization Firestone & Scholl (2016) Lupyan & Ward (2013) Thierry et al. (2009).
Historical Foundations: Linguistic Relativity as an Empirical Program
Twentieth-century debates concerning the relationship of language, culture, and thought motivated the empirical tradition commonly named linguistic relativity. The intellectual history is often organized around Edward Sapir and Benjamin Lee Whorf, but the empirical content is best recovered through specific predictions about classification, attention, reasoning, and memory. An association between a grammatical or lexical contrast and task performance is a research lead; demonstrating a causal role requires experiments or informative quasi-experiments. The enduring significance of linguistic relativity is that it turned universal assumptions about human conceptual organization into comparative questions. Its claims vary dramatically by domain, strength, and timescale, and the term should never be substituted for an operational mechanism Kay & Kempton (1984) Li & Gleitman (2002) Papafragou et al. (2008).
The experimental history of color supplied a tractable stimulus continuum with measurable physical and perceptual distances. Berlin and Kay's 1969 study argued that basic color vocabularies exhibit systematic constraints and developmental tendencies, countering the idea that all possible partitions of color are equally likely. Eleanor Rosch Heider's 1972 work on color naming and memory became a landmark in the study of candidate universal focal colors. Later studies reexamined cross-language variability and the relation of naming to recognition, making clear that the relevant dispute encompasses both constraints and local variation rather than a single yes-or-no test Berlin & Kay (1969) Heider (1972) Roberson et al. (2000).
Kay and Kempton's 1984 experiments sharpened the causal question. English and Tarahumara speakers were compared in color similarity judgments, and the authors proposed that participants could recruit lexical categories in making judgments. A follow-up procedure designed to block the relevant linguistic strategy altered the observed effect, suggesting task-sensitive mediation. This study became influential because it used a manipulation of the hypothesized mechanism instead of treating group differences as sufficient explanation. It also anticipated later work distinguishing stored category structure from online labeling and warned that the format of a question can determine whether language is engaged Kay & Kempton (1984) Cibelli et al. (2016) Winawer et al. (2007).
Universals, Local Regularities, and the Structure of the Debate
Human visual systems share physiological constraints, and all observers confront structured environments in which some contrasts are more useful than others. Such commonalities can generate similarities in category placement, discrimination, or preference even when languages have different vocabularies. The comparative question is therefore how much variation is left after stimulation, receptor-level limits, material contexts, and common communicative problems are considered. Berlin and Kay, later color-constancy research, and the study of ecological communicative utility all highlight mechanisms capable of shaping vocabularies without assuming an arbitrary mapping between words and experience Berlin & Kay (1969) Witzel & Gegenfurtner (2018) Gibson et al. (2017).
Variation can be substantial and orderly. Roberson, Davies, and Davidoff's work with Berinmo speakers reported category-linked differences in recognition and categorical judgments relative to English-speaking samples. Subsequent comparisons with another southern African population extended the claim that learned category boundaries contribute to cognition. These findings are informative because they involve populations that differ in color vocabularies and cultural contexts. They also illustrate interpretive difficulty: differences in formal education, instructions, familiarity with color chips, memory demands, language translation, and the mapping from name to button response must be evaluated as possible contributors Roberson et al. (2000) Roberson et al. (2005).
A defensible synthesis allows universal sensory constraints and culturally patterned behavior to coexist in a single explanatory account. The presence of a language effect in a particular category task establishes its conditional efficacy; the absence of the effect under different task conditions establishes a limit of that efficacy. Measurement therefore determines the reach of each conclusion. Research becomes more precise when it distinguishes a recurring cross-linguistic regularity from a within-participant intervention, and a spontaneous language strategy from durable learning-dependent changes in the weighting of information Kay & Kempton (1984) Cibelli et al. (2016) Firestone & Scholl (2016).
Theoretical Explanations: Attention, Categories, and Bayesian Inference
An attentional account proposes that learned labels make particular stimulus distinctions accessible and likely to be selected. A category is a useful compression of variation: it emphasizes differences relevant for communication while grouping physically diverse examples. When two alternatives straddle a category boundary, the task may exploit that partition and produce faster responses or larger estimated differences. The account predicts sensitivity to verbal interference, instructions, time pressure, and a person's currently active language. It also predicts that category effects may coexist with an underlying continuum of discrimination capacities Winawer et al. (2007) Sinkeviciute et al. (2024) Wakui et al. (2025).
A probabilistic-inference account instead specifies how noisy evidence and prior category knowledge are combined. Cibelli and colleagues modeled color reconstruction and cross-language discrimination as the consequences of uncertain perceptual information combined with category-consistent prior beliefs. This makes a mechanism testable: when sensory precision declines or memory uncertainty increases, prior categorical structure should have greater influence. The explanation is powerful because it produces quantitative predictions about reconstruction bias and can accommodate differences between short-delay memory and simultaneous comparison. It remains a model whose assumptions and parameter identification must be tested rather than an automatic demonstration of neural coding changes Cibelli et al. (2016) Roberson et al. (2000).
A stronger sensory-penetration hypothesis asserts that lexical information modifies perceptual representation before the decision stage. Electrophysiological and awareness-related experiments motivate investigation of such mechanisms, yet timing or performance alone does not identify causal direction. Firestone and Scholl's methodological critique identifies alternative explanations involving attention, judgments, memory, and task demands in claims of top-down perception. Evidence should thus be organized by where an intervention acts and by which later processes are held constant. Competing models may predict the same categorical advantage in one task and diverge in experiments that vary uncertainty, linguistic load, and response criterion independently Thierry et al. (2009) Lupyan & Ward (2013) Firestone & Scholl (2016).
Research Architecture: Domains, Questions, and Evidence Types
The first experimental family concerns color and other bounded stimulus continua, where researchers can independently measure physical distance, subjective similarity, categorical naming, and choice speed. A second family concerns complex scenes and events, where culture-linked allocation of attention may shape subsequent memory and explicit descriptions. A third concerns spatial and temporal reference, where language and literacy provide conventions for organizing relations that are not always directly available as perceptual properties. A fourth examines odor naming and hedonic evaluation, separating the capacity to verbalize a sensory difference from the sensory or affective value assigned to it Chua et al. (2005) Fuhrman & Boroditsky (2010) Majid & Burenhult (2014).
The underlying evidence is heterogeneous. Cross-sectional comparisons are useful for finding structured variability but leave alternative cultural covariates uncontrolled. Within-person language-context manipulations identify more local causal effects while potentially changing difficulty and the allocation of effort. Verbal-interference tasks can test recruitment of language, though the interfering task may impose general executive costs. Response-time, eye-tracking, electrophysiological, and memory paradigms each have distinctive construct-validity problems. A scientifically integrated account compares them without assigning the same explanatory power to every measurement Athanasopoulos (2009) Papafragou et al. (2008) Witzel & Gegenfurtner (2011).
The following chapters use these methodological distinctions to read foundational studies alongside independent failures to replicate specific signatures and more recent experiments. The evidential aim is a conditional account of where language and culture make measurable contributions, when those contributions disappear, and what experiments could locate them more exactly. A special emphasis falls on the dangers of treating a country as a homogeneous cognitive category or interpreting a late verbal response as a direct reading of sensory processing. The practical consequences extend to psychometric testing, education, cross-language interfaces, accessibility, and comparison between biological observers and artificial perceptual systems Henrich et al. (2010) Majid et al. (2018) Li et al. (2025).
Color Categorization: The Berinmo and Southern African Evidence
Color offers an unusually controlled test domain because researchers can present coordinates within a defined color space while measuring naming and discrimination. Yet equal distances in a display's numerical coordinates are not necessarily equal perceptual distances, and display calibration, luminance, saturation, adaptation, and illuminant assumptions remain essential. In the Berinmo experiments, Roberson, Davies, and Davidoff investigated color naming, memory, and category boundaries against a background of earlier universality claims. Their findings included category-linked response patterns that differed from English, showing that linguistic and local experience could affect experimental organization of the continuum Roberson et al. (2000) Heider (1972) Witzel & Gegenfurtner (2018).
The 2005 work of Roberson and colleagues extended the research to southern African speakers through similarity, short-term-memory, and learning tasks. A result observed across multiple kinds of judgment is more interesting than one restricted to the naming question used to define the groups. Nevertheless, memory and similarity may recruit category labels by different routes. Similarity can be judged through direct appearance, item identification, memory of comparison stimuli, or a strategy of matching names. Strong interpretation requires separating these routes experimentally and establishing the psychophysical distances of actual tested samples Roberson et al. (2005) Kay & Kempton (1984) Cibelli et al. (2016).
These findings do not imply that a language speaker cannot see a chromatic difference that lacks a separate word. They show how learned partitions of a continuous space can contribute to some measured choices and recollections. The distinction is central because categorical response structures emerge in many domains where the relevant sensory continuum persists. Replication needs to preserve the original boundary locations, evaluate group-specific naming maps, and compare within- and across-category pairs at matched distances. An experimental color chart is one representation of color experience, and transfer to material colors, objects, and variable illumination is a distinct research question Roberson et al. (2000) Witzel & Gegenfurtner (2018) Gibson et al. (2017).
Russian Blues: Lexical Boundaries and Verbal Interference
Winawer and colleagues compared Russian speakers, whose common basic terms distinguish lighter and darker blue, with English speakers, for whom the comparable region is normally organized under a broader term. A visual discrimination task required participants to identify a target shade under controlled comparisons. Russian participants showed a task-specific between-category advantage that was disrupted by verbal interference. The contrast between linguistic and nonlinguistic interference provided a mechanism probe: a speed benefit associated with stored lexical boundaries can depend on access to language during performance Winawer et al. (2007).
The experiment's inferential value lies in the interaction between group, boundary, and intervention. Merely being Russian or English predicts many differences in education, experience, and linguistic familiarity. Demonstrating that a verbal load alters the boundary advantage provides stronger evidence that linguistic resources are causally relevant at some stage. It does not establish that the earliest chromatic representation is reorganized or that every color discrimination task will yield the same effect. Attention, strategy selection, working-memory access, and response preparation remain candidate mediators and should be explicitly modeled Winawer et al. (2007) Firestone & Scholl (2016).
To evaluate the phenomenon across laboratories, investigators must determine actual category boundaries for individual participants and reproduce isoluminant or otherwise well-controlled stimuli. Simple translation of 'light blue' and 'dark blue' is insufficient because the semantic status of a basic color term may differ across communities. A complete analysis reports error and reaction time together, controls speed–accuracy trade-offs, compares verbal load with a matched general load, and avoids claiming an invariant effect when its strength depends on a particular choice architecture. Later bilingual experiments suggest that the active linguistic context is itself an experimental variable Athanasopoulos (2009) Sinkeviciute et al. (2024) Wakui et al. (2025).
Visual Hemifields and the Neural Interpretation of Language Effects
Gilbert and colleagues proposed that linguistic color-category advantages should be more prominent in the right visual field, whose initial sensory input projects principally to the left cerebral hemisphere. Their visual-search findings reported faster detection for differently named colors in the right visual field but not the left, motivating a language-lateralization interpretation. The result is an operational claim about a precise interaction among category status, target position, and reaction time. Treating that interaction as a general property of color vision requires independent replication using carefully measured colors and representative participants Gilbert et al. (2006).
Independent testing complicated that interpretation. Witzel and Gegenfurtner conducted multiple variants of the earlier paradigms involving 230 observers overall and found category-related response advantages in both visual fields without the reported lateralized pattern. Their analysis addressed stimulus calibration and the determination of genuine category boundaries. A later Czech conceptual replication by Štěpánková and Urbánek likewise reported a spatial pattern at odds with the original proposed asymmetry. These studies constrain the specific claim that an effect must occur predominantly in the right visual field, even while preserving the broader possibility of category-sensitive performance Witzel & Gegenfurtner (2011) Štěpánková & Urbánek (2023).
Behavioral hemifield asymmetry is an indirect neural measure. The side of a target influences orienting, response mapping, acuity, and the distribution of attention, while hemisphere-specific language systems interact across a connected brain. Accordingly, the validity of the proposed mechanism depends on converging evidence beyond a single laterality contrast. Pre-registered replications should compare models containing field, category, task, and participant-level slopes, include eye-position verification, and assess whether a category advantage remains after generic response and attention effects are modeled. A failed signature of lateralization is a meaningful scientific result, not a denial of all relationships between language and perceptual choice Gilbert et al. (2006) Witzel & Gegenfurtner (2011) Firestone & Scholl (2016).
Preattentive Responses, Awareness, and the Problem of Causality
Thierry and colleagues examined responses to color contrasts in speakers of languages that lexicalize the blue region differently and reported an event-related neural signature associated with language-specific category contrasts. This research expanded the evidence beyond button-press speed and made a valuable prediction about the timing of sensory-related processing. Neural timing, however, must be interpreted within the experimental design: differential responses can reflect stimulus history, category learning, automatic attention, expectancy, or comparison operations. A label-sensitive electrical response establishes that neural processing varies with a condition; it does not by itself isolate the first sensory transformation responsible for that difference Thierry et al. (2009) Firestone & Scholl (2016).
Lupyan and Ward demonstrated that congruent verbal labels can improve the entry of otherwise difficult visual objects into awareness under their particular detection conditions. The finding illustrates how a linguistic cue may set up feature templates and change access to a stimulus. The causal path could run through expectation and feature-based attention, or involve effects closer to sensory representation. Disentangling these requires controls that match cue information while varying linguistic form, and measurements that distinguish detection sensitivity from response bias. The awareness threshold is a property of both visual evidence and task architecture Lupyan & Ward (2013).
Across this literature, 'preattentive,' 'unconscious,' and 'perceptual' identify different claims. A neural response measured when participants are not explicitly categorizing color can still be modulated by an experimental context that makes categories meaningful. A discrimination accuracy gain could reflect more precise sensory estimates, more effective allocation of processing, or altered decision policy. The strongest experiments measure several of these candidate stages under the same manipulation and estimate the relative contribution of each. Methodological critiques of top-down perceptual effects reinforce the need to operationalize mechanisms rather than use an early timestamp as sufficient proof Thierry et al. (2009) Lupyan & Ward (2013) Firestone & Scholl (2016).
Bilingual Perception and the Active-Language Context
Bilingualism offers a particularly informative design because some aspects of learning history can be held constant within one observer. Athanasopoulos compared Greek–English bilinguals differing in exposure and proficiency with monolingual English speakers, exploiting Greek lexical distinctions between lighter and darker blue. The study found relationships among bilingual experience, semantic prototypes, and judgments of color similarity. This result suggests that linguistic experience and immersion may shape how a person groups and interprets stimuli, but it also shows that proficiency, cultural exposure, and label availability should be measured separately Athanasopoulos (2009).
In a 2024 study, Sinkeviciute and colleagues examined Lithuanian–Norwegian bilinguals and related monolingual comparison groups. The bilingual color-category effect differed with the language made active during the task. The finding extends the argument beyond fixed population traits: the same individual can recruit different linguistic resources across conditions. The effect under verbal interference in a particular language also cautions against treating 'verbal interference' as a uniform inhibitor of labeling across bilingual communities; the interference itself can activate a linguistic structure and influence performance. Design and analysis therefore require transparent treatment of active language and the content of interference Sinkeviciute et al. (2024).
Cross-language equivalence is especially demanding when linguistic labels differ in basicness or their distribution across contexts. Standardized color stimuli must be combined with an individualized naming assessment, proficiency measurement, language-use history, and comparable instructions. Within-person designs can establish a local causal contribution of activated language, but the intervention may also alter familiarity, confidence, executive workload, and expectancies. A strong program compares monolingual, balanced bilingual, and late-acquisition groups while retaining continuous individual characteristics rather than imposing a rigid two-culture classification Athanasopoulos (2009) Sinkeviciute et al. (2024) Henrich et al. (2010).
Ecological Utility, Communication, and Contemporary Replications
Color words are tools for identifying objects, coordinating action, and communicating useful distinctions. Gibson and colleagues compared naming across languages and examined how communication needs and perceptual structure jointly influence the distribution of color terms. Their work demonstrates that category structure can be explained in part by communicative utility in the environment, giving a causal research direction beyond abstract vocabulary counts. The history of exposure to natural objects, fabrics, manufactured materials, and shared practices may influence how frequently a contrast is named and which categories become stable Gibson et al. (2017) Witzel & Gegenfurtner (2018).
Contemporary work makes replication more discriminating. Wakui and colleagues' 2025 comparison of United Kingdom participants and Himba speakers used visual-search tasks designed to reduce label matching, finding effects linked to color-category boundaries. This extends the domain of investigation while placing the question of task-dependent categorical versus continuous networks at its center. A successful cross-linguistic difference in a controlled search task carries stronger evidential value than descriptions alone, yet it cannot answer every question concerning early sensory coding or everyday object recognition Wakui et al. (2025) Roberson et al. (2005).
The ecological account and the language-account are linked by feedback: environments make certain distinctions useful, communication stabilizes their labels, and labels may make a learned distinction more available during later choices. This relationship can explain both regularities across populations and locally different decision patterns. Experimental evidence must establish which part of the cycle is active in a given study. The need for careful control is heightened by the fact that 2023 conceptual replication and 2011 lateralization experiments produced outcomes contrary to a prominent neural-laterality narrative, whereas contemporary bilingual studies still find conditional language-related effects Gibson et al. (2017) Štěpánková & Urbánek (2023) Sinkeviciute et al. (2024). A complementary experiment directly compared simultaneous color matching with delayed matching in a sample of Persian speakers: Hasantash and Afraz found that vocabulary size and local color-name density correlated with memory performance but not simultaneous perceptual matching. The sample was modest and correlational, so the result sharpens the distinction between memory and sensation while leaving the causal direction of the association unresolved Hasantash & Afraz (2020).
Visual Scenes: Cultural Differences in Contextual Attention
Complex visual scenes require selecting information across many objects, surfaces, relations, and spatial scales. Nisbett and colleagues proposed a broad distinction between relatively analytic and holistic orientations in cognitive processing, relating it to the learned practices and social ecology of different populations. This framework generated testable predictions about what observers describe, remember, and attend to. Its value lies in motivating carefully specified tasks; the terms 'analytic' and 'holistic' designate patterns of performance rather than fixed attributes of every member of a cultural group Nisbett et al. (2001) Henrich et al. (2010).
Masuda and Nisbett compared Japanese and American participants viewing naturalistic underwater scenes and other images. Reports and recognition tasks suggested differential sensitivity to contextual information and to whether an object appeared against its original background. This evidence linked scene perception to later recognition and demonstrated that context can change the fate of an otherwise identical remembered object. Yet descriptions and recognition are downstream products of attention, familiarity, encoding, and memory. A stronger conclusion about visual processing requires measures obtained during inspection of the scene and interventions on environmental or task context Masuda & Nisbett (2001).
The scientific question becomes which features of scenes are prioritized under which goals. A cultural effect in a memory task may reflect environmental experience, conventions about salient information, or differences in descriptions encouraged by experimenters. The visual system processes both global spatial structure and local detail; what varies may be how a specific observer allocates limited resources. Researchers should compare the same scenes at equal exposure duration, track fixation histories, manipulate instructions, control visual acuity, and assess transfer to new materials. The relevant outcome is a reproducible relation among participant history, image statistics, and task, not a timeless national style of vision Chua et al. (2005) Li et al. (2025) Nisbett et al. (2001).
Eye Movements and the Selection of Visual Information
Chua, Boland, and Nisbett measured eye movements while American and Chinese observers inspected photographs with foregrounded objects and complex backgrounds. In their sample, Americans tended to fixate the focal object earlier and more often, whereas Chinese participants made more looks toward backgrounds. These observations help connect cultural group differences to the actual allocation of overt visual attention. Eye tracking is a valuable intermediary measure because it records behavior during viewing rather than only a final verbal or mnemonic report Chua et al. (2005).
A fixation is an observable sampling decision, and its interpretation depends on task, scene construction, central bias, salience, and the relationship between overt and covert attention. Participants may look at a background because it is relevant to recognition, because of a visual feature, or because the instructions make contextual relations important. Differences in fixation frequency should thus be assessed against the physical properties of the regions, viewing time, initial gaze position, and the opportunity to detect local details without eye movements. A methodologically strong design combines gaze data with later memory while avoiding an assumption that visual experience is exhausted by fixation location Chua et al. (2005) Masuda & Nisbett (2001).
Research on cultural context also teaches that the categories used to recruit participants need independent validation. A student who has spent years studying in a second country may occupy several communicative communities, and cultural practices may differ substantially within each recruitment group. Age, literacy, media exposure, drawing conventions, and educational specialization are measurable histories that may predict viewing strategies better than country labels alone. Multisite work should use individual-level predictors and treat group classifications as sampling variables, not as explanations in themselves Henrich et al. (2010) Li et al. (2025) Nisbett et al. (2001).
Scene Memory, Spatial Frequency, and Contemporary Evidence
Scene recognition integrates sensory encoding with short-term retention and a decision about a previously seen display. In a 2025 study, Li, Chien, Gutchess, and Sekuler compared young North American and East Asian adults on delayed matching of scenes whose spatial-frequency content was systematically filtered. Overall recognition accuracy did not differ between the groups, while patterns of confusions depended in part on which spatial-frequency information remained. This distinction is scientifically consequential: a null difference in mean success can coexist with structured differences in which information is weighted when uncertainty arises Li et al. (2025).
The study's design made its physical manipulations explicit and assessed visual characteristics of participants, yet generalization is limited by the sample composition, university context, and chosen image set. A cultural explanation of errors must also reckon with stimulus memorability and the statistical structure of spatial detail. An image of a room, an animal, and a landscape can differ in where diagnostically useful information lies. By analyzing error types rather than relying only on mean accuracy, investigators can distinguish a shift in the information used from a generalized advantage for one group Li et al. (2025) Chua et al. (2005).
This evidence connects the scene tradition to a broader principle: the organization of judgments is visible in conditional response distributions. Gaze duration, memory error direction, and sensitivity to image filtering provide complementary descriptions of performance, yet none alone establishes cultural causation. Replication should vary scene type, spatial scales, exposure conditions, educational settings, and individual visual histories. Especially important is comparing culture-sensitive outcomes with measurements of contrast sensitivity or acuity, so that high-level selection is not confused with differences in optical or physiological access to the stimuli Li et al. (2025) Masuda & Nisbett (2001) Henrich et al. (2010).
Motion Events: Linguistic Encoding and Task-Dependent Viewing
Languages differ in how they conventionally package motion events. Some descriptions favor verbs indicating manner of motion, while others emphasize path and trajectory through different lexical and grammatical resources. Gennari and colleagues compared English and Spanish speakers in nonlinguistic recognition and similarity tasks involving motion events. Their investigation linked lexicalization patterns to the kind of encoding required by an experimental task. Such findings establish an important bridge between the linguistic description of an event and subsequent decisions about its identity or resemblance Gennari et al. (2002).
Papafragou, Hulbert, and Trueswell compared English and Greek speakers while tracking eye movements during videos of motion. Participants showed language-related differences in attention while preparing verbal descriptions, including early in the event. When they freely inspected events for a memory task, eye movement patterns during the ongoing event were similar between language groups; certain differences emerged after the motion had stopped. The result challenges a simple prediction that habitual verb semantics always directs online visual attention. Instead, the active goal of describing or remembering appears to control when linguistic resources enter the task Papafragou et al. (2008).
This experimental contrast illustrates how research can separate habitual linguistic structure from the immediate demands of speaking. A group may differ robustly in how it describes a collision, ascent, or departure without continuously seeing a different trajectory. At the same time, a language-driven rehearsal strategy can alter memory and comparison even when initial gaze allocation is alike. Future tests should manipulate the probability of later speaking, compare unexpected recognition with planned description, and record both early attention and later responses. Only these distinctions establish the temporal location and functional significance of the observed differences Papafragou et al. (2008) Gennari et al. (2002) Firestone & Scholl (2016).
Spatial Reference Frames and the Environmental Organization of Judgment
Spatial judgments can depend on egocentric references, environmental landmarks, or conventional coordinate systems. Languages differ in the frequency and grammatical organization of expressions that specify relations using left and right, intrinsic object geometry, or bearings anchored to the surrounding world. Such differences make spatial reference an important test case for linguistic relativity because the experimenter can compare both spoken relations and nonverbal placement. Yet the ecological environment may simultaneously train orientation skills, making language use and practiced spatial navigation difficult to separate Li & Gleitman (2002) Henrich et al. (2010).
Li and Gleitman experimentally examined spatial reasoning in relation to the linguistic and environmental context of spatial problems. Their study underscored the importance of how objects, reference points, and room configurations are arranged when a supposedly nonverbal response is collected. A task with visible landmarks can recruit a distinct frame from a task with isolated objects, even with the same instructions. As a result, cross-language differences in spatial reconstruction require controls for available environmental cues, the orientation in which the participant is tested, and the response mappings learned during the session Li & Gleitman (2002).
A sound comparative program measures habitual spatial language, actual navigation practices, landmark availability, and responses to controlled rotations or transpositions. It distinguishes a participant's ability to use a spatial system from preference for that system under a specific task. The presence of a culturally typical coordinate strategy is compatible with broad flexibility, and flexibility itself is a legitimate outcome. An explanatory model should say when a frame is retrieved and how cue reliability, training history, and situational demands govern switching among frames Li & Gleitman (2002) Nisbett et al. (2001) Henrich et al. (2010).
Time, Reading Direction, and Cultural Spatial Mappings
Temporal order is often represented through spatial sequences, although the direction of the resulting arrangement may differ with writing practices and cultural conventions. Fuhrman and Boroditsky compared Hebrew and English speakers on tasks including the arrangement of temporal picture sequences and an implicit procedure intended to assess directional spatial–temporal associations. Their results linked language-community practices and directionality to some patterns of temporal representation. This is a different question from direct perception of the duration of an interval: the experiment concerns the spatial coding of temporal relations and the behavior of judgments about sequence Fuhrman & Boroditsky (2010).
Writing direction is a plausible mediator because habitual reading repeatedly aligns order and space. Nonetheless, language, script, schooling, handedness, interface exposure, and the design of response surfaces can covary. A left-to-right arrangement may be automatic for an experienced reader without constituting a permanent property of time experience. A compelling study would test multiple scripts, bidirectional readers, task frames, and within-person changes in current orientation. Distinguishing metaphor use from nonverbal spatial organization is necessary before extrapolating from a timeline arrangement to a general theory of temporal cognition Fuhrman & Boroditsky (2010).
The strongest inference from this domain concerns the use of culturally learned spatial codes during temporal judgments. Researchers should measure the stability of the code under nonlinguistic response conditions, the extent of its generalization beyond picture order, and its interaction with explicit temporal language. Results may bear on education, chart design, calendars, and interface directionality while leaving interval timing mechanisms as a separate empirical question. Such carefully bounded interpretation preserves the substantive achievement of cross-cultural time research and clarifies how linguistic and nonlinguistic practices can converge on a common response Fuhrman & Boroditsky (2010) Firestone & Scholl (2016) Majid et al. (2018).
Olfactory Vocabulary and the Diversity of Sensory Description
Human olfactory experience has often been discussed as difficult to describe, especially in English, where speakers regularly identify odor sources rather than using a rich set of abstract odor-quality terms. Majid and Burenhult challenged any presumed human universal of lexical limitation by studying Jahai, an Austroasiatic language with established odor terms. Their work showed that odor naming competence differs dramatically across language communities and invites a reexamination of what counts as successful sensory description. A limited English odor lexicon cannot be projected onto human perceptual language as a whole Majid & Burenhult (2014) Majid et al. (2018).
Wnuk and Majid studied Maniq, spoken by a small hunter-gatherer community in southern Thailand, and documented a structured set of odor terms through exemplar generation, similarity judgments, and evaluations. Their analysis linked semantic dimensions of smell to pleasantness and danger and showed that everyday ecological knowledge is embedded in odor vocabulary. The significance extends beyond lexical lists: a stable term can make an odor similarity relation more communicable and may guide what a listener expects to detect. Whether such linguistic organization directly improves raw olfactory thresholds requires a separate experiment with odor concentration and detection conditions controlled Wnuk & Majid (2014).
Comparative studies must distinguish odor identification, free naming, communication success, discrimination, recognition memory, and hedonic evaluation. A person can recognize two odors as different without supplying conventional names, or supply a highly specific label while showing average performance on low-level detection. Differences in odor familiarity may influence both semantic precision and judgments of pleasantness. The appropriate outcome therefore depends on the question: lexical expressibility is assessed through language tasks, while a claim about altered sensory discrimination needs comparable odorants, concentrations, sampling procedures, and response requirements Majid & Burenhult (2014) Wnuk & Majid (2014) Arshamian et al. (2022).
Cross-Modal Diversity and the Comparative Language of Sensation
Color is experimentally convenient, yet it is only one sensory domain. Majid and a large international team compared how languages encode perceptions drawn from multiple modalities and demonstrated substantial cross-language variation in what is easily expressed. This wider comparison matters because theories built around color may overgeneralize from a highly developed scientific measurement space. Smell, texture, taste, audition, and visual appearance differ in what can be presented repeatedly, what everyday practices make salient, and how easily listeners align their references to the same object Majid et al. (2018).
Cross-domain comparisons require domain-specific controls. A standardized color patch permits precise colorimetric reproduction, whereas a smell stimulus varies with concentration, airflow, prior exposure, and adaptation. Speech and environmental sound contain temporal structure that cannot be reduced to static resemblance. A tactile texture can be sampled actively, changing the stimulation through the observer's action. Thus a claim that language influences 'perception' must specify which modality, which measurement, and which kind of linguistic distinction is under investigation. The same category-learning mechanism may appear across tasks without generating an identical effect size or temporal profile Majid et al. (2018) Witzel & Gegenfurtner (2018) Wnuk & Majid (2014).
The comparative implication is methodological rather than merely taxonomic: cultural variation is most informative when it is studied against the actual opportunities for discrimination and action in a community. Differences in naming density may reflect communicative usefulness, familiarity, conceptual expertise, or learned perceptual contrasts. To isolate the contribution of lexical structure, research can teach new labels, vary immediate language activation, and compare objective identification with subjective similarity. Comparative domains should be treated as coordinated tests of a family of explanations, not as equivalent demonstrations of a single uniform cultural effect Gibson et al. (2017) Majid et al. (2018) Sinkeviciute et al. (2024).
Learning, Development, and the Dynamics of Category Acquisition
Language–perception relationships develop through learning and use. Children encounter recurring objects, caregivers' descriptions, cultural classifications, and tools that make particular perceptual contrasts meaningful. A color name may first be learned as a communicative convention and later become an efficient retrieval cue during comparison or memory. An adult who learns a second language may change the accessibility of category prototypes without erasing existing sensory discrimination. Longitudinal and training designs are therefore essential for understanding how a current category effect emerged rather than merely documenting its presence Athanasopoulos (2009) Roberson et al. (2005).
Cross-sectional contrasts between monolingual and bilingual groups cannot, by themselves, distinguish selection into a learning environment from learning caused by that environment. Within-person language-switch experiments estimate immediate activation effects, and repeated training with novel labels can ask whether a previously absent boundary advantage develops. Researchers should measure pretraining sensitivity, training history, retention, and transfer to unfamiliar colors or contexts. Ideally, new perceptual stimuli are assigned to lexical conditions independently of physical separability. Such designs can determine whether naming supports efficient comparison, stable memory, or durable changes in the interpretation of incoming evidence Sinkeviciute et al. (2024) Cibelli et al. (2016).
Developmental analyses must also contend with changing task comprehension, attention span, vision, schooling, and willingness to guess. A younger child may provide fewer distinct labels because words are unfamiliar while nonetheless distinguishing the relevant stimuli. Older participants may use strategic rehearsal that younger children cannot yet sustain. A well-powered study would employ nonverbal detection and matching measures as well as naming, calibrate difficulty for age, and model response bias. The central developmental question is how cultural learning changes the relationship among available sensory information, accessible categories, and the judgments people can reliably express Heider (1972) Frank et al. (2008) Henrich et al. (2010).
Artificial Perception and the Artificial Era: A Comparative Research Agenda
In the Artificial Era, perceptual judgments can be studied in both human observers and computational systems that convert images, sound, or other signals into classifications and descriptions. This comparison requires defining what each system receives, which prior categories it has learned, and what counts as a judgment. A vision model's color label, a language model's answer about an image, and a human subject's button press are observable outputs of very different architectures. Performance correspondence is scientifically useful when task input, scoring, uncertainty, and intervention are defined, rather than interpreted as proof of identical sensory experience Cibelli et al. (2016) Firestone & Scholl (2016).
A rigorous cross-system experiment might present color gradients or scenes under matched output demands and vary the words available in a prompt while keeping the image fixed. The measured quantity would be the change in classification or confidence caused by linguistic context, alongside analogous within-person human language manipulations. Such experiments can test representational dependence, sensitivity to naming, and robustness under paraphrase, but causal mechanisms must be studied separately within each architecture. An algorithm might exhibit a categorical boundary because its training labels encode a distinction, even where a human observer obtains the same response through attention and learned communication Gibson et al. (2017) Sinkeviciute et al. (2024).
This comparative program situates Psychology of Artificial within the broader scientific analysis without importing claims of equivalence from shared terminology. Its open questions concern the stability of category boundaries, transport to unfamiliar ecological settings, the causal effect of language on multimodal decisions, and criteria for functional comparability. Artificial Sapiens provides the authorial horizon in which nonbiological systems become independent objects of psychological inquiry. Empirical claims within this dossier remain tied to demonstrated observations and specified tests; theoretical possibilities concerning artificial psychological organization are treated as research propositions Majid et al. (2018) Witzel & Gegenfurtner (2018) Henrich et al. (2010).
Psychophysical Measurement and Decision Criteria
Perceptual judgments are experimentally accessible only through behavior or physiological signals, and measures differ in what they identify. A forced-choice discrimination task can estimate sensitivity to a physical stimulus difference, while a similarity scale blends remembered appearance and the participant's mapping of qualitative relations to numbers. Naming measures the agreement between a response and a lexicon; a same–different judgment includes a decision threshold. The most consequential design principle is to define the estimand before collecting data. If the question is whether language changes sensory sensitivity, a difference in response criterion or naming fluency cannot substitute for an independently estimated change in sensitivity Kay & Kempton (1984) Cibelli et al. (2016) Firestone & Scholl (2016).
Signal-detection reasoning offers a useful analytic distinction between overlap in internal evidence distributions and a criterion for selecting one response. Different categories can yield different response frequencies without greater separation of sensory evidence. Researchers may use forced-choice accuracy, fitted psychometric functions, and bias-sensitive signal-detection analyses, subject to the assumptions of their models. Reaction-time advantages likewise need evaluation against error rates, trial history, and response selection. An analysis that reports the full distribution of correct and incorrect responses, rather than only an average speed, is better positioned to discriminate attention, memory, and criterion accounts Winawer et al. (2007) Witzel & Gegenfurtner (2011) Štěpánková & Urbánek (2023).
Stimulus control is as important as statistical sophistication. In color experiments this means calibration and appropriate perceptual distance; in scene studies, matching visual structure; in smell research, known concentrations and adaptation procedures. Individual participants may differ in acuity, sensory history, or familiarity with a material. Researchers should compare within-category and between-category pairs without making the latter intrinsically easier, and test how estimates shift across task variants. These precautions transform a broad cultural claim into an identifiable experimental contrast and help determine whether a published discrepancy is theoretical or merely procedural Witzel & Gegenfurtner (2018) Li et al. (2025) Sorokowski et al. (2024).
Causal Identification: Language, Training, and Interference
Cross-population comparisons provide natural variation but rarely isolate language as the only cause of a difference. Languages co-occur with schooling, environments, cultural conventions, and response habits. Stronger inference is possible with randomized label training, instructed language activation in bilingual participants, and within-person tasks contrasting verbal and nonverbal interference. Each design identifies a different causal quantity: the immediate effect of a cue, the cumulative effect of learned categories, or the contribution of a strategy under task load. These must not be conflated, especially where findings are expressed in the common language of linguistic relativity Winawer et al. (2007) Athanasopoulos (2009) Sinkeviciute et al. (2024).
Verbal interference is informative when it selectively alters access to language, yet common interference tasks also consume working memory, motor preparation, and attention. A linguistic-load manipulation may activate the very language whose contribution is being measured, as contemporary bilingual evidence illustrates. An experiment should therefore use interference controls matched for difficulty, preregister the critical interactions, and examine whether the load changes baseline discrimination as well as the category-specific advantage. Interpretation rests on the difference between two controlled effects, not simply on the disappearance of a statistically significant comparison in one subgroup Sinkeviciute et al. (2024) Winawer et al. (2007).
Training studies can use arbitrary new categories and physically balanced stimuli, randomly assign label systems, and compare changes before and after exposure. The major threat is that the trained groups receive different attention or feedback, rather than only different linguistic structure. A label-training control can use the same stimulus statistics and rewards but provide an equivalent nonlinguistic code. Testing transfer and retention evaluates whether the effect is a short-lived strategy or a learned reorganization of accessible information. Quantitative models are particularly useful when they specify how evidence precision and category priors jointly predict judgment changes Cibelli et al. (2016) Gibson et al. (2017) Lupyan & Ward (2013).
Sampling, Translation, and Cross-Cultural Measurement Invariance
Comparative psychology has repeatedly relied on narrow samples drawn from highly educated industrialized populations. Henrich, Heine, and Norenzayan showed why generalizing from such samples to human cognition requires independent evidence, including for tasks described as fundamental or culturally neutral. Cross-cultural perceptual research shares this problem in both directions: findings from a university sample should not stand for all members of a nation, and a small field sample should not stand for an entire linguistic tradition. Recruitment, exposure history, literacy, and task familiarity are part of the evidence, not administrative details Henrich et al. (2010).
Instructions must convey the same decision across languages. Translating an English term may import a category contrast absent from a local language, or a familiar local term may carry contextual meanings not represented in English. Back-translation is useful but cannot alone establish functional equivalence. Researchers should pilot instructions with local collaborators, use comprehension checks that do not presuppose target categories, record the source language of response coding, and document how experimenters negotiated examples. The central criterion is that the same experimental question and response opportunity are understood by each participant Majid et al. (2018) Roberson et al. (2005) Wnuk & Majid (2014).
Measurement invariance also concerns apparatus and sampling ecology. A monitor differs by spectral output and viewing conditions; odor delivery differs by storage and local temperature; visual-scene images contain culturally uneven familiarity. Portable protocols should report calibration, exclusion rules, participant characteristics, missingness, and item-level variation. Hierarchical analysis can model participants and stimuli as sources of uncertainty rather than averaging away meaningful heterogeneity. Multisite replication gains value when it adapts procedures to preserve construct equivalence while keeping physical evidence and decision requirements comparable Li et al. (2025) Sorokowski et al. (2024) Wakui et al. (2025).
Critical Synthesis: Where the Evidence Converges and Diverges
The color literature offers converging evidence that lexical categories can influence some behavioral judgments, including memory, similarity, and discrimination under specified tasks. Category-boundary advantages in Russian, Berinmo, Himba, and bilingual comparisons indicate that learned distinctions matter. The causal force and locus of those effects remain more constrained: verbal interference, active language, and task settings can change their manifestation. Independent experiments have not consistently reproduced the stronger claim of an exclusively right-visual-field advantage. Thus the supported synthesis is a domain- and task-conditioned influence on judgments, with neural localization and sensory penetration requiring additional evidence Roberson et al. (2000) Winawer et al. (2007) Witzel & Gegenfurtner (2011) Wakui et al. (2025) Sinkeviciute et al. (2024).
The scene and event literature locates much of the effect in information selection and task engagement. Cultural comparisons of eye movements and contextual recognition show patterned differences in viewing or memory, while Greek–English motion research found language-related viewing differences predominantly when participants prepared to describe events. Spatial and temporal studies similarly emphasize learned relational codes that are recruited under particular arrangements. These findings are scientifically important as explanations of selection, reference, and memory without requiring a global alteration of sensory capacity Chua et al. (2005) Masuda & Nisbett (2001) Papafragou et al. (2008) Fuhrman & Boroditsky (2010).
Olfaction reveals a complementary combination of linguistic diversity and common sensory constraints. Rich odor lexicons in Jahai and Maniq undermine claims that smell intrinsically resists precise language, whereas cross-cultural odor ranking demonstrates substantial shared structure for selected molecular stimuli. Ecological familiarity and individual variation remain explanatory factors. Across domains, the recurring pattern is an interaction among stimulus properties, learning history, accessible concepts, and current decisions. No single study supports a universal assertion that language always determines perception, and the literature is most useful when each result is interpreted at the level actually measured Majid & Burenhult (2014) Wnuk & Majid (2014) Arshamian et al. (2022) Sorokowski et al. (2024).
Open Problems and a Cumulative Research Program
The unresolved problem is to locate linguistic and cultural influences along the causal path from stimulation to observable response. The same reaction-time effect may arise from attentional deployment, perceptual estimation, lexical retrieval, decision policy, or motor mapping. Research should combine calibration, eye tracking, psychophysical thresholds, confidence, reaction-time distributions, and neural measures under the same randomized manipulation. Models that specify different predictions for these channels can be compared prospectively. This approach is especially necessary where strong statements about neural lateralization or sensory penetration have outpaced replication of the critical behavioral interaction Firestone & Scholl (2016) Witzel & Gegenfurtner (2011) Cibelli et al. (2016).
A second problem concerns development and generalization. How quickly can new labels change decisions? Does an acquired boundary persist when labels are unavailable? Does it transfer from chips to objects under natural illumination, or from laboratory odorants to ecological mixtures? Longitudinal and field experiments can estimate the relative contributions of language learning, exposure, expertise, and communicative utility. Cross-language comparisons should include variation within communities and repeated measurements across years, rather than treating linguistic category inventories as static descriptions of human minds Athanasopoulos (2009) Gibson et al. (2017) Majid et al. (2018).
A third problem concerns reproducible comparison across different kinds of observers and technological environments. Online experiments, multilingual communication, and artificial perceptual models open new testing opportunities while introducing differences in display calibration, sampling, and training data. Multisite protocols should preserve deidentified analysis data where ethical and legal arrangements permit, publish code and stimulus specifications, and distinguish discovered patterns from preregistered predictions. For Artificial Era research, a functional comparison of human and artificial judgment must report system input, language conditions, outputs, scoring rules, and known architectural differences. This is an agenda for falsifiable interdisciplinary experiments rather than an inference of shared phenomenology Henrich et al. (2010) Wakui et al. (2025) Li et al. (2025).
Conclusion
Culture and language shape perceptual judgments through demonstrable interactions among learned categories, available words, attention, memory, communicative demands, and ecological experience. The most developed evidence comes from color experiments, where categorical distinctions can alter response speed or remembered similarity and where verbal interference and bilingual language activation provide informative causal manipulations. Their effects are conditional on procedure, population, and task. The failure of specific visual-field signatures to generalize is part of this evidential structure and narrows the interpretation of broad neurocognitive hypotheses without negating established category effects Winawer et al. (2007) Witzel & Gegenfurtner (2011) Sinkeviciute et al. (2024) Wakui et al. (2025).
Research on scenes, motion, space, time, and odor establishes that the language–perception relationship varies by domain. Scene differences often concern selection of contextual information and memory; motion studies reveal task-sensitive attention during description; spatial and temporal experiments show culturally practiced reference systems; olfactory comparisons demonstrate both dramatic linguistic diversity and common chemical constraints on pleasantness. These findings converge on an account that takes shared biological capacities and historically acquired cognitive tools as joint determinants of behavior. Scientific claims acquire precision by specifying their dependent measure, intervention, and population Papafragou et al. (2008) Fuhrman & Boroditsky (2010) Majid et al. (2018) Arshamian et al. (2022).
The cumulative scientific task is to estimate where and how these contributions operate, using psychophysical control, representative sampling, translation validation, causal interventions, and models that distinguish sensitivity from criterion. This monograph establishes a research framework for linking original cross-cultural experiments to rigorous measurement and explicitly testable future work. In the historical horizon From Homo to Artificial, the same disciplinary rigor enables comparison of human and artificial classifications while respecting the mechanisms and evidence of each kind of system. Culture and language belong within the scientific explanation of perceptual judgment because their effects are experimentally investigable, variable, and consequential Cibelli et al. (2016) Firestone & Scholl (2016) Henrich et al. (2010).
References
- Arshamian, A., Gerkin, R. C., Kruspe, N., Wnuk, E., Floyd, S., O'Meara, C., Garrido Rodriguez, G., Lundström, J. N., Mainland, J. D., & Majid, A. (2022). The perception of odor pleasantness is shared across cultures. Current Biology, 32(9), 2061–2066.e3. https://doi.org/10.1016/j.cub.2022.02.062
- Athanasopoulos, P. (2009). Cognitive representation of colour in bilinguals: The case of Greek blues. Bilingualism: Language and Cognition, 12(1), 83–95. https://doi.org/10.1017/S136672890800388X
- Berlin, B., & Kay, P. (1969). Basic color terms: Their universality and evolution. University of California Press. https://books.google.com/books?id=xatoCJbhifkC
- Chua, H. F., Boland, J. E., & Nisbett, R. E. (2005). Cultural variation in eye movements during scene perception. Proceedings of the National Academy of Sciences, 102(35), 12629–12633. https://doi.org/10.1073/pnas.0506162102
- Cibelli, E., Xu, Y., Austerweil, J. L., Griffiths, T. L., & Regier, T. (2016). The Sapir-Whorf hypothesis and probabilistic inference: Evidence from the domain of color. PLOS ONE, 11(7), Article e0158725. https://doi.org/10.1371/journal.pone.0158725
- Firestone, C., & Scholl, B. J. (2016). Cognition does not affect perception: Evaluating the evidence for ‘top-down’ effects. Behavioral and Brain Sciences, 39, Article e229. https://doi.org/10.1017/S0140525X15000965
- Frank, M. C., Everett, D. L., Fedorenko, E., & Gibson, E. (2008). Number as a cognitive technology: Evidence from Pirahã language and cognition. Cognition, 108(3), 819–824. https://doi.org/10.1016/j.cognition.2008.04.007
- Fuhrman, O., & Boroditsky, L. (2010). Cross-cultural differences in mental representations of time: Evidence from an implicit nonlinguistic task. Cognitive Science, 34(8), 1430–1451. https://doi.org/10.1111/j.1551-6709.2010.01105.x
- Gennari, S. P., Sloman, S. A., Malt, B. C., & Fitch, W. T. (2002). Motion events in language and cognition. Cognition, 83(1), 49–79. https://doi.org/10.1016/S0010-0277(01)00166-4
- Gibson, E., Futrell, R., Jara-Ettinger, J., Mahowald, K., Bergen, L., Ratnasingam, S., Gibson, M., Piantadosi, S. T., & Conway, B. R. (2017). Color naming across languages reflects color use. Proceedings of the National Academy of Sciences, 114(40), 10785–10790. https://doi.org/10.1073/pnas.1619666114
- Gilbert, A. L., Regier, T., Kay, P., & Ivry, R. B. (2006). Whorf hypothesis is supported in the right visual field but not the left. Proceedings of the National Academy of Sciences, 103(2), 489–494. https://doi.org/10.1073/pnas.0509868103
- Hasantash, M., & Afraz, A. (2020). Richer color vocabulary is associated with better color memory but not color perception. Proceedings of the National Academy of Sciences, 117(49), 31046–31052. https://doi.org/10.1073/pnas.2001946117
- Heider, E. R. (1972). Universals in color naming and memory. Journal of Experimental Psychology, 93(1), 10–20. https://doi.org/10.1037/h0032606
- Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2–3), 61–83. https://doi.org/10.1017/S0140525X0999152X
- Kay, P., & Kempton, W. (1984). What is the Sapir–Whorf hypothesis? American Anthropologist, 86(1), 65–79. https://doi.org/10.1525/aa.1984.86.1.02a00050
- Li, H., Chien, J., Gutchess, A., & Sekuler, R. (2025). Visual short-term memory, culture, and image structure. Attention, Perception, & Psychophysics, 87, 1630–1641. https://doi.org/10.3758/s13414-025-03094-7
- Li, P., & Gleitman, L. (2002). Turning the tables: Language and spatial reasoning. Cognition, 83(3), 265–294. https://doi.org/10.1016/S0010-0277(02)00009-4
- Lupyan, G., & Ward, E. J. (2013). Language can boost otherwise unseen objects into visual awareness. Proceedings of the National Academy of Sciences, 110(35), 14196–14201. https://doi.org/10.1073/pnas.1303312110
- Majid, A., & Burenhult, N. (2014). Odors are expressible in language, as long as you speak the right language. Cognition, 130(2), 266–270. https://doi.org/10.1016/j.cognition.2013.11.004
- Majid, A., Roberts, S. G., Cilissen, L., Emmorey, K., Nicodemus, B., O'Grady, L., Woll, B., LeLan, B., de Sousa, H., Cansler, B. L., Shayan, S., de Vos, C., Senft, G., Enfield, N. J., Razak, R. A., Fedden, S., Tufvesson, S., Dingemanse, M., Ozturk, O., … Levinson, S. C. (2018). Differential coding of perception in the world's languages. Proceedings of the National Academy of Sciences, 115(45), 11369–11376. https://doi.org/10.1073/pnas.1720419115
- Masuda, T., & Nisbett, R. E. (2001). Attending holistically versus analytically: Comparing the context sensitivity of Japanese and Americans. Journal of Personality and Social Psychology, 81(5), 922–934. https://doi.org/10.1037/0022-3514.81.5.922
- Nisbett, R. E., Peng, K., Choi, I., & Norenzayan, A. (2001). Culture and systems of thought: Holistic versus analytic cognition. Psychological Review, 108(2), 291–310. https://doi.org/10.1037/0033-295X.108.2.291
- Papafragou, A., Hulbert, J., & Trueswell, J. (2008). Does language guide event perception? Evidence from eye movements. Cognition, 108(1), 155–184. https://doi.org/10.1016/j.cognition.2008.02.007
- Roberson, D., Davidoff, J., Davies, I. R. L., & Shapiro, L. R. (2005). Color categories: Evidence for the cultural relativity hypothesis. Cognitive Psychology, 50(4), 378–411. https://doi.org/10.1016/j.cogpsych.2004.10.001
- Roberson, D., Davies, I., & Davidoff, J. (2000). Color categories are not universal: Replications and new evidence from a stone-age culture. Journal of Experimental Psychology: General, 129(3), 369–398. https://doi.org/10.1037/0096-3445.129.3.369
- Sinkeviciute, A., Mayor, J., Vulchanova, M. D., & Kartushina, N. (2024). Active language modulates color perception in bilinguals. Language Learning, 74, 40–71. https://doi.org/10.1111/lang.12645
- Sorokowski, P., Misiak, M., Roberts, S. C., Kowal, M., Butovskaya, M., Omar-Fauzee, M. S., Huanca, T., & Sorokowska, A. (2024). Is the perception of odour pleasantness shared across cultures and ecological conditions? Evidence from Amazonia, East Africa, New Guinea, Malaysia and Poland. Biology Letters, 20(6), Article 20240120. https://doi.org/10.1098/rsbl.2024.0120
- Štěpánková, L., & Urbánek, T. (2023). Colour categorization and its effect on perception: A conceptual replication. Journal of Psycholinguistic Research, 52(1), 1–16. https://doi.org/10.1007/s10936-021-09791-2
- Thierry, G., Athanasopoulos, P., Wiggett, A., Dering, B., & Kuipers, J.-R. (2009). Unconscious effects of language-specific terminology on preattentive color perception. Proceedings of the National Academy of Sciences, 106(11), 4567–4570. https://doi.org/10.1073/pnas.0811155106
- Wakui, E., Mylonas, D., Caparos, S., & Davidoff, J. (2025). Categorical color perception shown in a cross-lingual comparison of visual search. Color Research & Application, 50(4), 301–313. https://doi.org/10.1002/col.22964
- Winawer, J., Witthoft, N., Frank, M. C., Wu, L., Wade, A. R., & Boroditsky, L. (2007). Russian blues reveal effects of language on color discrimination. Proceedings of the National Academy of Sciences, 104(19), 7780–7785. https://doi.org/10.1073/pnas.0701644104
- Witzel, C., & Gegenfurtner, K. R. (2011). Is there a lateralized category effect for color? Journal of Vision, 11(12), Article 16. https://doi.org/10.1167/11.12.16
- Witzel, C., & Gegenfurtner, K. R. (2018). Color perception: Objects, constancy, and categories. Annual Review of Vision Science, 4, 475–499. https://doi.org/10.1146/annurev-vision-091517-034231
- Wnuk, E., & Majid, A. (2014). Revisiting the limits of language: The odor lexicon of Maniq. Cognition, 131(1), 125–138. https://doi.org/10.1016/j.cognition.2013.12.008
Social Communication and the Practical Value of Labels
Perceptual terms enable joint reference: two speakers can coordinate attention to a particular hue, odor, scene relation, or event feature without transmitting an exhaustive physical description. A lexical category can be advantageous even where perceptual differences already exist, because it reduces ambiguity and memory burden. Color naming research reveals systematic pressures from communicative usefulness, while experiments on language activation show that terms available at a given moment can change how a task is completed. These are mechanisms of coordination and cognitive economy whose usefulness can be measured directly Gibson et al. (2017) Sinkeviciute et al. (2024).
Communicative efficiency can also feed back into perceptual priorities. A community may repeatedly distinguish ripe from unripe materials, safe from dangerous smells, or relevant from irrelevant motion patterns. Such recurrence makes certain features well practiced and certain labels reliable indicators of meaning. An ecological explanation should therefore include both the frequency of encounters and the frequency of successful communicative exchanges. It should not infer cultural values directly from a lexical inventory; the inventory must be connected to documented practices, contexts of use, and individual experience Wnuk & Majid (2014) Gibson et al. (2017) Majid et al. (2018).
An informative experimental design separates individual classification from collaborative communication. Participants first make private judgments, then solve a reference task with partners using controlled vocabularies, and finally complete a delayed nonverbal discrimination or memory task. Changing the usefulness of a label tests whether communication experience shifts its later availability. A control group with equivalent sensory practice but no shared naming can isolate the incremental linguistic mechanism. This links field observations about culturally situated vocabulary to experimentally defined, observable causal pathways Kay & Kempton (1984) Gibson et al. (2017) Cibelli et al. (2016).