On the Declared and the Determinable

A cosmetic formulation is required by regulation to declare its ingredients. In the European Union, this declaration takes the form of an INCI list — International Nomenclature of Cosmetic Ingredients — printed on the packaging in descending order of concentration. The requirement exists to enable informed choice and to support post-market safety surveillance.

It does not exist to enable structural analysis.

The distinction matters. A declaration is a list of what is present. A structural description is an account of how those components behave in relation to one another, at what concentrations, under what conditions, and with what architectural consequence. These are different informational objects. Regulation requires the first. The second does not exist as a public document for any commercially available formulation.

This is not a gap that will be closed by better regulation. It reflects a structural feature of the cosmetic industry: formulation architecture is proprietary. The INCI list is the only surface that manufacturers are required to make visible. Everything beneath it — concentration, pH, emulsion type, processing parameters — remains, with rare exceptions, undisclosed.

Consider what the INCI list does and does not contain.

It contains component identity — the names of ingredients in a standardized nomenclature that allows cross-referential lookup against safety databases and published literature. This is not nothing. It allows the identification of known sensitizers, membrane-disruptive agents, and pH-active compounds. It provides a first-order structural signal.

What it does not contain is concentration. The descending order requirement holds only above one percent by weight. Below that threshold, ingredients may be listed in any order. For a formulation where ten or fifteen components fall below one percent — actives, preservatives, chelating agents, fragrance components — the list provides no quantitative information. The same INCI sequence could represent a trace inclusion or a functionally significant loading.

It does not contain pH. This is the most significant single omission for formulations targeting compromised skin barriers. pH is not a component — it is a system property, a consequence of the interaction between buffer systems, acidic or alkaline components, and the aqueous phase. It directly regulates serine protease activity in the stratum corneum. A formulation operating outside the functional pH corridor of 4.5 to 5.0 actively destabilizes compromised barriers, independent of any other property. This parameter is not declared.

It does not contain emulsion architecture. Whether a system is a conventional oil-in-water emulsion, a lamellar liquid crystal, a microemulsion, or a Pickering system is not determinable from the ingredient list alone. The same components can be processed into fundamentally different structural architectures with profoundly different barrier interactions. The list is silent on which architecture was achieved.

It does not contain process parameters. Emulsification temperature, homogenization energy, cooling rate, and filling conditions determine the physical state of the final product in ways that component identity cannot predict. Two formulations with identical INCI lists, produced under different process conditions, may behave as structurally distinct systems at the skin interface.

The consequence of these absences is not that external analysis is impossible. It is that external analysis is structurally constrained — and the degree of constraint varies systematically with what is being asked.

The question of whether a formulation contains a known membrane-disruptive agent is largely answerable from the INCI list. The question of whether the formulation is physiologically coherent as a system is not.

This distinction is not a matter of analytical ambition. It is a matter of what the available evidence permits. A framework that evaluates structural coherence without acknowledging this constraint is not rigorous — it is producing verdicts that the underlying data cannot support.

The honest analytical position is not to refuse evaluation in the absence of complete data. It is to be explicit about what the available data permits, to evaluate within those limits, and to state the limits as part of the analytical output. A partial assessment, clearly delimited, is more scientifically defensible than a complete assessment built on inferred data.

There is a further complication that is less often acknowledged: the declared formulation is not always a stable object.

INCI lists are published across multiple channels — product packaging, brand websites, retailer pages, pharmacy databases, regulatory submissions. These versions are not always identical. Minor reformulations may be implemented without public announcement. Regional variants may carry different compositions under the same product name. Translation between INCI nomenclature and national pharmacopoeia designations introduces ambiguity in pharmacy-distributed products. The result is that the declared composition of a single product may exist in multiple non-identical versions simultaneously.

For external analysis, this creates a specific problem: the object of analysis is indeterminate. An assessment conducted against one version of the declared composition may not apply to another version of the same product. This is not an edge case — it is a routine condition of formulation analysis based on public data.

The appropriate response is not to ignore the discrepancy or to choose a version without acknowledgment. It is to document the discrepancy, specify which version was used as the basis for assessment, and note the effect of the discrepancy on analytical confidence.

None of this constitutes a criticism of the cosmetic industry as such. Formulation architecture is legitimately proprietary. The competitive value of a formulation lies precisely in the specific combination of components, concentrations, and process parameters that produces a particular result. Mandatory disclosure of this information would eliminate the structural basis for product differentiation. The regulatory requirement is calibrated to consumer protection, not to external scientific analysis.

The point is not that more should be disclosed. The point is that what is disclosed defines the outer boundary of what can be determined. External formulation analysis operates within that boundary — and must be honest about where the boundary is.

This honesty has a practical implication. An analytical institution that issues verdicts on structural coherence without specifying the data conditions under which those verdicts were reached is not producing scientific output. It is producing opinion formatted as analysis. The formatting does not change what it is.

The analytical contribution is not the verdict alone. It is the verdict together with an explicit account of what the verdict is based on, what it cannot be based on, and what would need to be known to reach a more complete determination. That account is itself a scientific product — and in a field where the data conditions are structurally constrained, it may be the more important one.

The declared and the determinable are not the same thing. The distance between them is not a failure of analysis. It is the condition under which analysis in this field operates. Acknowledging that condition precisely is where rigorous external assessment begins.

Notes on constrained systems ·