The Italian colour lexicon in Tuscany: elicited lists, cognitive salience, and semantic maps of colour terms

The study addresses how Tuscan Italian speakers conceptualize and lexicalize colours, testing universalist versus relativist accounts of basic colour categories (BCCs) and their naming. Building on the idea that BCTs reflect universal perceptual constraints but show language-specific variation, the authors examine the salience and structure of Italian colour terms, with special interest in multiple basic ‘blue’ terms reported for Italian and potential emerging basic terms. They employ free-list elicitation to quantify psychological salience (frequency and position in lists) and to map conceptual relationships among terms, aiming to estimate the Tuscan BCT inventory, probe the status of the three ‘Tuscan blues’ (blu, azzurro, celeste), and gauge the salience of high-frequency non-BCTs.

Prior work proposes that languages evolve a limited set of basic colour categories named by basic colour terms (Berlin & Kay), with later accounts allowing cross-linguistic variation shaped by perceptual and cultural constraints (e.g., Regier et al.). Non-basic terms can become basic over time (Kay & McDaniel). Salience measures from free-listing (frequency, earliness) distinguish BCTs from non-BCTs (Corbett & Davies), and the Cognitive Salience Index (S) combines frequency and mean position (Sutrop). Adjacency-based semantic mapping using MDS and clustering reveals listing strategies and conceptual chunks (Uusküla & Bimler). For Italian, psycholinguistic studies have identified counterparts of the 10 B&K BCTs and evidence for more than one basic ‘blue’ term, with regional variation: blu and azzurro in Veneto/Trento, celeste and blu in Alghero-Catalan, and triple ‘blues’ in Tuscany. Italian exhibits rich derivational and compounding mechanisms that refine colour naming, influenced by cultural history, fashion, and environmental referents (e.g., sea and sky).

Participants: 89 native Italian speakers (68 female), aged 20–31 (mean 23.2 ± 3.0), born and residing in Tuscany. Procedure: A 5-minute free-listing task was administered online (Zoom/Meet). Participants received an Excel sheet via Google Drive and were instructed (in Italian) to "Please write as many colours as you know." The experimenter timed the task and announced elapsed time each minute. Data processing: Morphologically modified forms were counted separately; spelling regularized; typos corrected; certain variant forms standardized (e.g., carta da zucchero variants collapsed); duplicates within a list removed, retaining first occurrence. Measures: For each term, frequency across lists (F) and mean list position (mP) were computed, with ranks R_F and R_mP. A Zipf-style popularity function was analyzed by plotting log10(number of participants listing the term) vs log10(rank by frequency), fitting a trilinear (double-power law) model to identify segments (kernel vs more specific lexicon). Analyses were restricted to frequent colour terms listed by at least 5 respondents (N_FCTS=96). Cognitive Salience Index S was computed as S = F/(N × mP), yielding R_S. Semantic map: Conceptual adjacency was estimated via inter-term separations in list order: SEP_pij = |n_pi - n_pj|, with ADJ(i,j) = exp[−(∑ SEP_pij)/c] over participants for whom both terms co-occurred. MDS (SPSS PROXSCAL) with non-metric ordinal transforms produced 2D and 3D solutions (S-Stress reported); hierarchical clustering (Ward’s method) on MDS coordinates (MATLAB) identified clusters. Ethics: Approved by the University of Florence ethics committee; informed consent obtained.

• Productivity and inventory size: 2675 items produced; 337 unique terms. Individual lists ranged 15–58 items; mean list length 30.06, indicating rich chromonymic competence among Tuscan speakers.
• Cognitive salience: The top S-ranked terms (1–13) comprised 10 counterparts of Berlin & Kay’s BCTs and three basic ‘blue’ terms (blu, azzurro, celeste) for Tuscan speakers. Primary chromatic BCTs led S: giallo (S=0.1966), rosso (0.1937), blu (0.1640), verde (0.1580), followed by nero (0.0989). Bianco showed relatively lower salience (S=0.0768; R_S=9). Viola was highly frequent/salient (R_S=7), reflecting cultural prominence.
• Emerging and salient non-BCTs: Fucsia had high salience (S=0.0568; R_S=14), close to celeste (R_S=13), suggesting an emerging BCT. Other salient non-BCTs included lilla (R_S=15), verde acqua (R_S=16), ocra (R_S=17), magenta (R_S=18), beige (R_S=19), bordeaux (R_S=20), indaco (R_S=21), oro (R_S=22), argento (R_S=23), porpora (R_S=24), turchese (R_S=25).
• Zipf-function: A trisegment function fit (overall R²≈0.9) with slopes: limb 1 ≈ −0.0295 (10 BCTs at very high popularity), limb 2 ≈ −0.9409 (ranks 11–24: includes grigio and the Tuscan azzurro and celeste alongside popular non-BCTs such as fucsia, lilla, beige, verde acqua, bordeaux, ocra, magenta, oro, argento), limb 3 ≈ −1.2114 (ranks 25–96: lower-popularity non-BCTs).
• Mean position (earliness): Primary BCTs had the highest priority (lowest mP). Anomalies included bianco and marrone occurring later than expected; fucsia occurred earlier than marrone.
• Semantic maps: MDS 2D (S-Stress=0.113) and 3D (S-Stress=0.0685) on S>0.01 terms (N_HS=32) revealed organization along three criteria: (1) a salience/priority gradient (D1), (2) linguistic economy/word length (D2), and (3) chromatic content vs lightness/desaturation (D3). Clusters showed chunking of related terms (e.g., green derivatives with turchese), adjacency of near-synonyms (amaranto–bordeaux–porpora), and positions reflecting Tuscan triple blues, with blu most inclusive and azzurro/celeste nearer rosa/viola; ciano gravitated toward azzurro/celeste.
• Areas of lexical refinement: High salience and variety in BLUE–GREEN (indaco, turchese, ciano), RED–PINK–PURPLE (fucsia, lilla, magenta, bordeaux, porpora, amaranto), and at the YELLOW periphery (ocra, beige), plus metallic sheens (oro, argento).

Findings support an augmented Tuscan Italian BCT inventory that includes three basic ‘blue’ terms (blu, azzurro, celeste), aligning with prior psycholinguistic evidence and suggesting language- and region-specific elaboration driven by environmental salience (sea/sky) and cultural history (textiles, fashion). The high salience and borderline status of fucsia indicate it may be emerging as a culturally basic term, consistent with its distinct niche between PINK and PURPLE and widespread usage in fashion and everyday discourse. The rich inventory and long free lists reflect strong cultural competence and communicative needs for nuanced shade distinctions, facilitated by Italian morphological and compounding strategies. Semantic mapping shows structured recall driven by salience, brevity, and chromatic content vs lightness, with consistent chunking of related hues and derived forms. Cross-language comparisons reveal similar refinement patterns (e.g., turquoise/teal, lilac, beige/ochre), highlighting universal communicative and perceptual pressures modulated by culture. The observed anomalies (lower salience of bianco among young speakers; high salience of viola) underscore sociocultural influences on salience beyond universal hierarchies.

The study delineates the Tuscan Italian colour lexicon using free-list elicitation, cognitive salience metrics, Zipf analysis, and semantic mapping. It establishes an augmented set of 13 BCTs for Tuscan speakers, including blu, azzurro, and celeste, and identifies fucsia as a likely emerging basic term among younger speakers. Beyond this core, numerous hyponyms, modified and compounded forms demonstrate high lexical productivity and communicative precision for colour nuances. The semantic maps reveal systematic organization by salience, linguistic economy, and chromatic/lightness attributes. These results illustrate how universal perceptual constraints interact with culture-specific histories to shape colour lexicons and suggest continued convergence and refinement under globalization and cultural transmission. Future research should extend analyses across Italian regions and age groups, integrate corpus-based frequency measures and behavioural naming metrics, and examine gender-balanced samples to refine estimates of basicness and salience.

• Regional and age scope: Data were collected only in Tuscany from young adults; results may not generalize across Italian regions or age cohorts.
• Salience measures: Frequency and cognitive salience do not always produce a sharp BCT/non-BCT boundary; complementary corpus and behavioural measures (consistency, consensus, response times) are needed.
• Gender imbalance: Sample skewed female (68 women, 21 men); potential gender differences in colour vocabulary warrant balanced sampling.
• Possible inter-generational differences: Emerging and loanword terms may be more prevalent among younger speakers; comparisons with older cohorts are needed.