An odour can be described in a file, but experiencing it still requires matter. Starting from this distinction, the paper traces four layers of constraint: physical, systemic, socio-cognitive and regulatory-institutional. It separates recognising an odour from reproducing it and examines the strongest alternative to physical distribution: transmitting a recipe for local production. The argument applies to the current technical regime and specifies what would overturn it. It then compares four possible homes for shared vocabularies, protocols and formats. None emerges as the clear winner; ORIZOMIA remains a proposal for shared content and rules of structural impartiality, with its institutional home, funding and adoption still open.

Every unit of fragrance requires molecules, energy and compliance; improving production cannot make those costs disappear. The paper distinguishes this material floor from the costs of ambiguous briefs, incompatible documents and repeated checks across the supply chain. It develops a cost taxonomy, examines the accounts of two fragrance composition houses and uses allergen labelling to show where individual efficiency depends on collective coordination. Its scope is the production of fragrant materials, with branded marketing and distribution considered to mark that boundary. Shared infrastructure could reduce informational friction, but the scale of potential savings remains unmeasured in this sector.

Ambiguous information can set off a costly sequence: conflicting interpretations, repeated exchanges, longer coordination and indirect expense. This analysis identifies where shared linguistic and measurement infrastructure might interrupt that sequence without changing physical or regulatory constraints. The paper examines brief development, compliance and audit in detail, alongside eight other areas of coordination. It also asks when standardisation costs more than it saves, which organisational alternatives deserve consideration and which creative ambiguities should remain open. Its contribution is a testable framework of mechanisms, indicators and adoption conditions, supported by indirect evidence; it does not estimate savings or demonstrate them empirically.

A technical vocabulary can acquire the force of a decision when a valid description becomes sufficient grounds for action. The paper calls this recurring pattern functional coincidence when it involves no autonomous deliberative step. It distinguishes semantic, epistemic and decision-making authority. Its practical test asks whether decision-makers retain a documented capacity to reject the descriptive content, rather than simply counting how often they follow it. European harmonised standards provide a documented case. Structural impartiality names the opposing arrangement: visible, autonomous deliberation and no preferential treatment based on position. The argument is conditional, limited to highly formalised systems, with ORIZOMIA considered as a design proposal rather than an operating institution.

A dictionary alone cannot connect the different jobs fragrance language performs. ORIZOMIA proposes an eight-level architecture linking the form and quality of experience with vocabulary, use, ambiguity, regulatory constraints, relations and panel calibration. At its centre, five axes describe experiential form separately from quality regions anchored to reference substances. Domain Profiles adapt the framework to particular settings, while bridge relations govern movement between them and shared governance controls revision. The paper sets out how the architecture, including its core axes, could be challenged and changed. Operational applications and AI integration remain proposals; the framework has yet to be applied to a real system.

What would another fragrance classification add to the systems already in use? The paper compares six established approaches, from commercial wheels and house vocabularies to IFRA categories, sensory profiles and ISO methods, across seven dimensions of coverage and governance. Each fulfils a different mandate. Within the comparison grid, none combines all the features proposed for ORIZOMIA, including explicit ambiguity management, codified uses and shared interoperability. Yet the grid partly reflects ORIZOMIA’s own architecture, and the sector’s need for that combination remains unproven. The proposed relationship is coexistence through mappings, with ORIZOMIA assessed as a design rather than a functioning competitor.

Fragrance classifications carry assumptions about language, identity, bodies, ritual and the work behind materials and sensory panels. Drawing on linguistic, sociological and sensory research, the paper examines how these practices shape the categories that technical systems often leave unexplained. It keeps the claim bounded: cultural variation is especially relevant to words, uses and market categories, while evidence suggests substantial agreement across cultures on the pleasantness of individual molecules. Edwards’ replacement of ‘Oriental’ with ‘Amber’ illustrates how classification can make its assumptions explicit. The proposed documentation within ORIZOMIA remains untested, as does the prediction of more systematic descriptor disagreement between social groups than within them.

Today’s fragrance families have histories that their familiar names can conceal. This genealogy follows classifications from pre-industrial manuals through synthetic chemistry and named accords to regulatory and methodological institutions. It shows how commercial prototypes, stable ingredient palettes and practical mandates helped categories take hold, while distinguishing documented developments from historical hypotheses. ‘Chypre’, for example, predates its twentieth-century prototype; the suggested link between ‘Oriental’ and Rimmel’s earlier imagery remains inferential. Existing classifications retain validity within their purposes. The historical record does not yet establish which families capture perceptual structure and which chiefly stabilise names, leaving ORIZOMIA with an inheritance to examine rather than a natural order to reproduce.

Competing fragrance businesses already share professional references, but moving between their vocabularies can still require interpretation. The paper asks whether mappings and provenance rules could be maintained as a common resource without sharing proprietary product knowledge. It uses the knowledge-commons framework to examine how this shared resource could be established and maintained, drawing conditional lessons from neighbouring sensory fields and comparing possible institutional hosts. No institutional model clearly emerges as the best choice. Crucially, the coordination gap itself remains a hypothesis requiring primary research, separate from the question of whether ORIZOMIA could address it. The contribution is a set of criteria for testing the need and designing shared coordination alongside existing institutions and classifications.

A fragrance brief works partly because it leaves room for interpretation. The paper proposes four forms of indeterminacy—lexical, pragmatic, contextual and intentional—and distinguishes them by what would resolve each one: examples, questions about audience and use, background knowledge or work on samples. This gives perfumers a way to decide which uncertainties to reduce and which to preserve for creative development. Optional documentation in ORIZOMIA’s proposed use and ambiguity layers would record those choices. The examples are illustrative, and the typology awaits testing on real briefs. Its scope is Western perfumery with time for iterative development; other traditions and faster workflows would require recalibration.

The same fragrance moves through different languages in research, marketing, compliance, training and sales. The paper builds a business case for a company vocabulary that connects these registers while preserving their purposes and could reduce what it calls a ‘translation tax’. It examines benefits across functions alongside maintenance costs, language-model alternatives and the risks of integrating AI without a shared linguistic reference. The case draws on established research and analogies, with no primary evidence from perfumery and no guaranteed return. A proposed audit of clarification requests lets a house test the argument before investing. A sector reference such as ORIZOMIA would require separate funding, governance and adoption.

An olfactory AI model can inherit disagreements embedded in its training labels. For technical leaders, this paper follows the chain from data coding and panel calibration to model performance and operational decisions. It proposes four necessary but insufficient conditions for industrial use: structured coding, calibrated panels, data governance and documented datasets and models. The Principal Odor Map provides an important limit: broad descriptions of single molecules can work with curated catalogue labels. The stronger case for coding concerns finer distinctions, house vocabularies, transfer between panels and mixtures. These remain working hypotheses; a proposed two-panel test could challenge the central claim, and ORIZOMIA is one possible framework rather than an established requirement.

Bringing AI into fragrance compliance means working across rules with different purposes and responsibilities. This paper maps European cosmetics and allergen requirements, IFRA Standards and the AI Act, distinguishing legal duties, contractual obligations and prudent practice. It separates deterministic compliance calculations from tasks that may benefit from models, including document extraction and descriptive drafting. Its proposed framework connects system records to the cosmetic safety dossier through formal validation, records protected against alteration, qualified supervision and clear responsibility. A staged adoption programme begins with substance identification and data governance. The framework draws on analogies with pharmaceuticals and medical devices; its effectiveness remains to be tested against established manual workflows.

Olfactory standards already govern sensory methods, ingredient safety, environmental measurement and data management. The difficulty arises when a term or dataset moves between them. Drawing on official documents, this paper examines the work of connecting these systems that falls outside their individual mandates: shared meanings, context of use, recognised ambiguities, links between regulatory regimes and documentation for AI. Its comparison identifies five points of connection that no individual mandate covers through a framework explicitly shaped by the proposed ORIZOMIA architecture. ORIZOMIA is presented as a developing means of connecting existing systems. Whether these interfaces create sufficient demand for a shared infrastructure remains an open question, to be examined through documented cases and a pilot.

Beyond a controlled sensory profile, what still needs to be investigated when a fragrance is experienced in a classroom, clinic or workplace? Building on existing sensory standards, this paper proposes an exploratory framework for comparing laboratory assessments with experience in shared environments. It treats context as a variable to document and test, with domain profiles for education, healthcare and work. The proposed university pilot involves adults and uses the same stimuli and measures in a laboratory and classrooms. Classrooms are randomly allocated, with at least two per arm; a power calculation sets the required sample size. The framework offers a research programme rather than findings about an odour’s effects. The regulatory conditions for ambient diffusion must also be established before the pilot can begin.

A molecule alone cannot account for what someone smells. E = F(M, B, P, C) proposes a relational account of olfactory experience, bringing together the material stimulus, biological constraints, a person’s prior psychological and memory profile, and context. Its central hypothesis is non-additivity: the effect of one component depends on the others. The paper translates these components into observable variables and nine conditional hypotheses, then explores perfume, flavour, and social and natural odours without collecting new data. A taxonomy of modelling failures links missing components to patterns in prediction errors. The resulting experimental programme specifies a decisive challenge: an additive model predicting new observations as well as a model with interactions chosen in advance.

Listing the influences on smell does not specify how they combine. This paper formalises a non-additivity criterion for E = F(M, B, P, C): a mixed derivative, or an interaction contrast for discrete components, must be non-zero for a pair selected in advance on the basis of a documented mechanism. It examines how that criterion depends on measurement scale, including which ordinal interactions survive monotonic transformations. Four implementations are distinguished from Partial Information Decomposition, a diagnostic whose synergy does not establish non-additivity. The paper sets out identification conditions and a programme of empirical testing. Evidence exists for specific molecule–receptor genotype interactions; interactions across all four components await data collected together from the same participants.

Which limits of AI in perfumery can research reduce, and which require a different response? This paper separates constraints arising from current data and architectures, structural difficulties that would persist within realistically attainable datasets, and institutional arrangements that can change through decisions. It examines mixtures, individual variability, explanation, safety constraints and attribution, stating the assumptions behind each classification and, where possible, evidence that could change it. A further discussion treats effects on creativity, formula diversity and tacit knowledge as hypotheses. Four research priorities follow: public mixture datasets, independent validation, IFRA constraints within optimisation, and longitudinal assessment on skin. The classification is intended to guide research expectations rather than declare permanent technological boundaries.

What makes a perfume the same work when its formula changes, and what makes its creation an authored act? This essay tests Goodman’s autographic–allographic distinction against reformulation, imitation and perfume’s performance on a body. It proposes an ephemeral, embodied category of work whose identity is established through shared olfactory comparison with a reference. Its discussion of AI separates a limiting case, where a generated formula is adopted unchanged, from documented practice in which a perfumer selects proposals. In the latter, authorship resides in judgement informed by compositional experience. These are philosophical arguments about identity and intention; favourable preference tests do not establish general equivalence with senior perfumers’ work.

A perfumer learns to judge a fragrance through its development on skin; generating more formulae cannot compress that experience. Drawing on Nietzsche, Anders and Connerton, this essay examines three conditions absent from statistical generation: selective forgetting that enables creative rupture, embodied competence acquired over time, and practices transmitted through shared physical presence. Its concern is the formation of future judgement. The thesis is conditional: if selecting generated proposals replaces learning to compose, present quality could remain stable while the competence sustaining future quality erodes. The essay reports no study of current perfumery training. It proposes interviews and comparative composition tasks that could test the mechanism and challenge its account.

The origin of an ingredient and the origin of a formula are different kinds of connection. Through Peirce, Barthes, Walton and Connerton, this essay distinguishes the perfume’s liquid, which retains a causal link to the materials used, from the decisions that shape its recipe. A perfumer’s choices can depend on particular materials they have smelled; a model generating from formula collections reproduces patterns in recipes without inheriting that encounter. The proposed term ‘icon of indices’ names this design relation. The argument applies to generation without measured batch information and changes when such information, or a perfumer’s material judgement, enters the process. It proposes criteria for assessing a recipe that openly declares this absence.

Does an evocative brief make a perfume less beautiful, or change how its beauty is judged? Revisiting Roudnitska’s use of Kant, this essay argues that free and adherent beauty distinguish modes of judgement rather than fixed categories of perfume. Knowing the brief supplies a concept of what the fragrance should be; judging against that concept does not make its beauty inferior. The proposed practical test asks whether declaring the brief and evaluation criterion improves agreement between assessors. The account remains a philosophical argument with an untested prediction, while the suggested influence of Roudnitska on industry distinctions is explicitly conjectural. The olfactory sublime marks a further territory beyond this framework.

Professional perfumery turns immediate experience into names, families and scales. This essay reads the tension between experience and classification through Nietzsche’s Apollonian and Dionysian pair, as an analogy proposed by the author. Difficulty naming odours is located within Western olfactory culture, not attributed universally to the sense itself. The argument is that classification remains productive when it leaves room for experience beyond the existing repertoire. A genealogical inquiry then examines the preference for natural over synthetic, French over functional, and niche over mass-market perfumery. Their histories remain partly conjectural. The resulting questions concern training beyond classificatory precision and evaluation that recognises inherited hierarchies as conventions rather than intrinsic aesthetic facts.

A perfume can be called ‘artistic’, ‘niche’ and ‘fine fragrance’, yet each label answers a different question. This essay separates application, market position and artistic status, following the terms through industry history, acquisitions, brand documents and theories of art and judgement. Fine fragrance identifies perfume for the skin; niche has shifted from restricted distribution towards production, audience or price; artistic claims a status that institutions may recognise in individual works. The essay argues that treating the labels as synonyms lets each borrow authority from the others. Documented cases illustrate the mechanism without measuring its prevalence. A proposed corpus analysis would test when the labels accumulate and whose interests that accumulation serves.

Fragrance families such as chypre, fougère and woody organise an entire industry, but their usefulness does not establish that they describe natural kinds. This paper examines the similarities that hold a family together, the gradual boundaries that classification makes discrete, and the different meanings the same words carry across trades. Its central concern is how conventions are revised: established classifications do change, yet the decisions remain with small committees or individual authors, without public criteria or a forum for their users. The proposal is to retain these families as useful, revisable tools and give those who rely on them a role in revising them.

A smell involves a physical stimulus, a source and what we say it means. The olfactory lekton names this third layer: meaning sustained through shared use, distinct from both the molecule and an individual’s perception. The paper traces the gradual passage from physiological constraints to perceptual construction and inherited cultural meanings. Its distinctive concern is the fragility of the samples on which olfactory words are learned: they evaporate, change through reformulation or survive in archives beyond everyday use. It proposes comparisons that could test whether this instability changes classifications, and explains why preserving a sample alone cannot preserve the meaning a community gives it.

On a product label, a technical definition and a buyer’s understanding can give the same word different conditions of truth. This paper follows that divergence through alcohol-free, fragrance-free, hypoallergenic and eau de parfum, distinguishing anchors that are missing, negative or tied to a different property. Its sharpest example is unscented: smelling may correctly confirm the absence of an odour while giving no assurance about masking agents or allergens. Other gaps arise when a threshold has never been fixed or a historical formula can no longer be checked. The proposed taxonomy records each term’s technical and commercial uses separately, identifies whether a claim concerns perception or composition, and dates changes to its reference.

Musk, sandalwood and chypre retain their names while the materials those names identify change, sometimes sharing a label for decades. This paper treats such words as positions occupied by successive or overlapping referents. Changes of reference occur in many fields; perfumery’s distinctive problem is the missing record. No shared authority preserves distinct terms for the earlier meanings of family and material names. A faithful taxonomy would therefore record historical meanings alongside current ones, including their periods of coexistence. The cases illustrate this pattern without establishing its prevalence, and show how an undated vocabulary can pass obsolete associations into models trained on fragrance descriptions.

Learning to make perfume requires a trained body as well as chemical knowledge. Perfumery schools have already formalised much of that learning, notably through Jean Carles’s method and their collections of reference materials. This paper asks what a school collection cannot provide on its own: references shared across institutions, calibration of individual differences, and comparison between cultural vocabularies. Drawing on literature and public programme descriptions, it proposes ways to teach these dimensions while retaining the tacit part of expertise. Wider access would depend on affordable teaching and public references. Without that infrastructure, proprietary digital tools could supply the vocabulary that becomes the profession’s default standard.

Control over smell often resides in its surrounding infrastructure: secret formulas, computational maps and their datasets, raw-material supply chains, and accounts of origin. The paper examines this distribution through a comparison with colour, where a publicly governed reference system remains available beneath proprietary vocabularies. Olfaction has standards, classifications and open datasets, but lacks an equivalent common reference layer; no present notation predicts the perceived smell of a mixture. The proposed counterweight is a knowledge commons built from measurement protocols, declared provenance and vocabularies anchored to samples. The concentration of control is presented as a diagnosis to test, with indicators that could challenge it, rather than an established measurement of the sector.

Music, colour, wine and radiology offer different ways of making expert perception shareable. This paper distinguishes notation that specifies an intended perceptual result from vocabularies tied to common samples, which let experts describe and dispute what they perceive. Perfumery remains limited on both fronts. Predicting a mixture’s smell from its formula is still an empirical challenge; building shared descriptive references also depends on institutional choices and on accords that lack ready-made counterparts in nature. Canonical samples could support a more public vocabulary. They would reduce reliance on tacit knowledge without transferring the experience itself, and broaden access only if learning that vocabulary becomes affordable.

Luxury perfume can be compared for pleasure, longevity and projection, as enthusiast communities and makers of alternatives already demonstrate. Rarity, authorship and heritage are harder for buyers to verify, even when they are described fluently. This paper locates the price premium chiefly in those unverifiable attributes, rather than in an inability to talk about smell. It examines how restricted comparison can soften price competition, including the role of L’Oréal v Bellure, and considers whether independent consumer communities could build shared procedures for comparison. The argument assumes neither universal cost ratios nor uniform accuracy among alternatives; any effect of price on perfume pleasure remains a hypothesis requiring direct evidence.

A fragrance pyramid may list many notes, while mixture research suggests that even trained people usually identify only three or four components at once. This paper asks whether that limit carries over to finished perfumes, distinguishing learned perceptual units from ingredients and from the contribution of unnamed materials to the whole. A study of 216 fragrance listings, sampled opportunistically and coded by one person, and a comparison across sources document descriptions rather than perception. From that distinction it develops criteria for comparable accounts: recognised units, separate composition data, timed protocols, identified assessors and measured agreement. Seven descriptive approaches are compared, and three testable predictions set out how the proposed transfer from laboratory mixtures could be assessed.

ORZ separates the data describing a perfume, the rule placing those data in time, and the figure generated from them. Its tables record ordinal intensity at ten fixed times, preserve the origin of each row and distinguish an unrecorded value from zero. Declared composition remains separate from perception; time-based derivations follow an explicitly named and versioned rule and are shown as calculated results. Making earlier plotting rules explicit exposed errors and discrepancies that their figures concealed. The notation checks structure rather than meaning. Two exercises using published descriptions explore possible disagreement without measuring smell, and the paper defines a protocol for the study of agreement between human assessors still needed to assess its use.

A model learns what its training labels record; whether those labels predict perception must be measured. This paper examines that distinction in molecular catalogues and fragrance pyramids, where declarations, editorial choices, votes and calculations may occupy the same column. Panel validation is encouraging for individual molecules, but the reviewed literature offers no equivalent validation for perfume pyramids. Comparisons across public descriptions reveal disagreement without establishing whether its source is editorial choice or noisy perception. The proposal makes the target, each value’s provenance, timing protocol, assessor agreement and vocabulary anchored to reference samples explicit. Three testable predictions and a minimum panel dataset define what is needed to evaluate the approach.