AUTHORS: [Author Name]
DATE: 2026-10-01
ABSTRACT:
Psychiatry depends heavily on systems of classification for describing mental disorders, organizing clinical observations, conducting research, communicating diagnoses, and planning treatment. Yet the systems used to classify psychiatric phenomena differ substantially in purpose, structure, assumptions, and units of analysis. This whitepaper proposes a taxonomy of psychiatric taxonomies: a framework for classifying the classification systems themselves. It compares categorical diagnostic systems such as DSM-5-TR and ICD-11 with dimensional and research-oriented frameworks such as the Hierarchical Taxonomy of Psychopathology (HiTOP) and the National Institute of Mental Health Research Domain Criteria (RDoC). The analysis suggests that psychiatric taxonomies can be distinguished along several major dimensions: purpose, unit of classification, structural organization, evidentiary basis, level of analysis, and relationship between normal and abnormal functioning. Rather than treating one taxonomy as universally definitive, psychiatry may be better understood as operating through several partially overlapping classification systems optimized for different scientific and clinical purposes.
- INTRODUCTION
Taxonomy is the systematic organization of phenomena into classes and relationships. In psychiatry, taxonomy becomes especially important because many psychiatric conditions are defined primarily through patterns of symptoms, behavior, cognition, emotional regulation, functional impairment, developmental history, and clinical observation rather than through a single laboratory marker.
The American Psychiatric Association’s DSM-5-TR organizes recognized mental disorders into a diagnostic classification and provides symptom-based diagnostic criteria. DSM-5-TR was published in 2022 and continues to receive approved updates.
The World Health Organization’s ICD-11 provides another major classification system. Its mental, behavioural and neurodevelopmental disorders framework forms part of the international system used for disease classification and health reporting. WHO released its detailed ICD-11 Clinical Descriptions and Diagnostic Requirements for these disorders in 2024.
These systems demonstrate that psychiatry is not merely concerned with identifying disorders. It must also determine how disorders should be grouped, separated, named, measured, and related to one another.
This produces a second-order problem: if psychiatry contains multiple taxonomies, how should those taxonomies themselves be classified?
The central research question is therefore:
How can psychiatric classification systems themselves be systematically classified according to their purpose, structure, unit of analysis, evidentiary foundations, and conception of psychopathology?
- PROBLEM STATEMENT
Psychiatric classification faces several conceptual difficulties.
First, categorical classifications divide psychopathology into named disorders. A person either satisfies a defined diagnostic threshold or does not, although severity specifications and dimensional measurements may also be incorporated.
Second, symptoms frequently cross diagnostic boundaries. Anxiety, sleep disturbance, impaired concentration, emotional dysregulation, anhedonia, cognitive changes, and other features may occur in multiple disorders.
Third, patients can satisfy criteria for more than one diagnosis. This phenomenon of diagnostic comorbidity raises questions about whether apparently separate categories sometimes represent overlapping manifestations of broader psychopathological dimensions.
Fourth, psychiatric classifications serve different purposes. A taxonomy created to support clinical communication may not be identical to one created to discover biological mechanisms. A statistical classification used for public-health reporting may require different properties from an experimental framework used in neuroscience.
Fifth, classifications may operate at very different levels. One system may classify syndromes, another traits, another symptoms, and another neurobehavioral processes.
Consequently, arguments about which psychiatric taxonomy is “correct” may occasionally compare systems designed to solve different problems.
A taxonomy of taxonomies can make these differences explicit.
- PROPOSED SOLUTION
This paper proposes classifying psychiatric taxonomies using six dimensions.
A. PURPOSE
Psychiatric taxonomies can first be classified according to their principal purpose.
Clinical taxonomies:
Designed to assist clinicians in recognizing, communicating, and describing mental disorders.
Statistical and administrative taxonomies:
Designed to standardize disease reporting, coding, epidemiology, and health-system information.
Research taxonomies:
Designed primarily to support scientific investigation of psychological, behavioral, biological, developmental, or neurological processes.
Transdiagnostic taxonomies:
Designed to examine phenomena that extend across conventional diagnostic boundaries.
These purposes can overlap.
B. UNIT OF CLASSIFICATION
A second distinction concerns what is actually being classified.
Possible units include:
- Disorders
- Syndromes
- Symptom clusters
- Individual symptoms
- Personality or psychopathology traits
- Behavioral dimensions
- Cognitive processes
- Neurobiological systems
- Developmental processes
- Functional impairment
DSM and ICD primarily organize recognized disorders and syndromes. DSM describes recognized mental disorders and specifies criteria clinicians use when evaluating diagnoses.
HiTOP instead organizes psychopathology hierarchically, ranging from broad spectra to narrower components, traits, symptoms, and maladaptive behaviors.
RDoC operates differently again. It investigates dimensions of human functioning using constructs that may be studied through such levels as genes, molecules, circuits, physiology, behavior, and self-report.
C. STRUCTURE
Psychiatric taxonomies may also be classified structurally.
Categorical:
Phenomena are divided into distinct diagnostic classes.
Dimensional:
Psychopathology is represented along continua of severity or functioning.
Hierarchical:
Narrow phenomena are nested within broader dimensions.
Network-oriented:
Symptoms or behaviors may be modeled according to their relationships with one another rather than assumed to arise exclusively from a single latent disorder.
Hybrid:
Categorical and dimensional elements are combined.
HiTOP provides a particularly clear example of a dimensional and hierarchical model. Its organization ranges from broad super-spectra and spectra through narrower subfactors, symptom components, traits, individual symptoms, and maladaptive behaviors.
D. SOURCE OF CLASSIFICATION
Taxonomies can further be distinguished by how their structures are constructed.
Consensus-informed systems rely substantially on expert review, accumulated clinical research, field testing, and formal revision processes.
Empirically derived systems attempt to identify structures appearing in observed data, such as correlations among symptoms and traits.
Mechanism-oriented frameworks classify psychological or biological functions that investigators believe may contribute to psychopathology.
These approaches are not mutually exclusive. Modern classification systems usually combine empirical evidence, expert interpretation, practical constraints, and historical development.
E. LEVEL OF EXPLANATION
Psychiatric phenomena can be classified at multiple explanatory levels:
SOCIAL AND ENVIRONMENTAL
↓
DEVELOPMENTAL
↓
BEHAVIORAL
↓
PSYCHOLOGICAL
↓
COGNITIVE
↓
PHYSIOLOGICAL
↓
NEURAL-CIRCUIT
↓
CELLULAR
↓
MOLECULAR
↓
GENETIC
A psychiatric taxonomy may emphasize one level or integrate several.
RDoC deliberately encourages research using multiple units of analysis and examines constructs across the range from normal to abnormal functioning. NIMH states explicitly that RDoC is a research framework rather than a diagnostic guide intended to replace current clinical diagnostic systems.
F. BOUNDARY BETWEEN NORMALITY AND DISORDER
Perhaps the most fundamental taxonomic distinction concerns whether psychopathology is conceptualized primarily as separate disease categories or as extreme, impaired, or dysfunctional positions along dimensions that exist throughout the population.
A categorical model emphasizes boundaries.
A dimensional model emphasizes degree.
A hierarchical model asks at what level of abstraction psychopathology should be described.
A biological or functional framework asks which underlying systems produce the observed psychological and behavioral phenomena.
These represent different answers to different questions.
- IMPLEMENTATION
The proposed meta-taxonomy can be represented as a classification matrix.
SYSTEM: DSM-5-TR
PRIMARY PURPOSE: Clinical diagnosis and research communication
PRIMARY UNIT: Mental disorder
STRUCTURE: Predominantly categorical with dimensional elements
LEVEL: Symptoms, behavior, functioning, clinical characteristics
USE: Clinical diagnosis, research, communication
SYSTEM: ICD-11
PRIMARY PURPOSE: International clinical classification and health reporting
PRIMARY UNIT: Disease or disorder
STRUCTURE: Categorical classification with dimensional components in selected areas
LEVEL: Clinical syndrome and functioning
USE: Clinical practice, epidemiology, statistics, health systems
SYSTEM: HiTOP
PRIMARY PURPOSE: Empirical organization of psychopathology
PRIMARY UNIT: Dimensions, traits, spectra, symptoms
STRUCTURE: Dimensional and hierarchical
LEVEL: Observable psychopathology
USE: Research, assessment, development of transdiagnostic approaches
SYSTEM: RDoC
PRIMARY PURPOSE: Mechanism-oriented research
PRIMARY UNIT: Neurobehavioral constructs
STRUCTURE: Dimensional matrix
LEVEL: Genes through behavior and self-report
USE: Experimental and translational research
This implementation avoids forcing all systems onto a single spectrum from “better” to “worse.” Instead, each taxonomy is described according to the problem it was designed to address.
- RESULTS AND DISCUSSION
Applying the proposed framework reveals that psychiatric classification is not one taxonomy but a collection of taxonomic strategies.
DSM-5-TR primarily answers:
Which recognized diagnostic description best characterizes this clinical presentation?
ICD-11 additionally addresses:
How can health conditions be represented consistently across international clinical and statistical systems?
HiTOP asks:
What hierarchical and dimensional structure emerges when psychopathology is organized according to patterns among symptoms and traits?
RDoC asks:
Which fundamental behavioral and biological systems can be studied across traditional diagnostic boundaries?
This distinction is important because the systems should not always be interpreted as direct competitors.
RDoC, for example, explicitly states that it is not intended as a clinical diagnostic guide.
Likewise, HiTOP represents psychopathology dimensionally and hierarchically rather than simply providing another list of conventional diagnostic categories.
The taxonomy-of-taxonomies approach therefore produces a broader conceptual model:
PSYCHIATRIC TAXONOMY
├── Diagnostic taxonomy
│ ├── DSM
│ └── ICD
│
├── Dimensional psychopathology taxonomy
│ └── HiTOP
│
├── Functional and mechanistic research taxonomy
│ └── RDoC
│
└── Future integrative taxonomies
├── Clinical phenotype
├── Symptom dimensions
├── Development
├── Environment
├── Cognition
├── Behavior
├── Neurobiology
└── Genetics
The broader implication is that psychiatry can itself be viewed partly as a science of classification under uncertainty.
Classification does not merely label observations. It determines what researchers count, what clinicians compare, what disorders epidemiologists track, which participants enter studies, and which relationships become scientifically visible.
For this reason, taxonomy is not peripheral to psychiatry. It is one of the field’s organizing problems.
A mature psychiatric classification architecture may eventually operate as a layered system rather than requiring a single taxonomy to perform every task.
For example:
Layer 1: presenting symptoms
Layer 2: severity and impairment
Layer 3: dimensional psychopathology
Layer 4: diagnostic syndrome
Layer 5: developmental trajectory
Layer 6: environmental context
Layer 7: cognitive and behavioral mechanisms
Layer 8: biological mechanisms
Such an architecture would allow researchers and clinicians to describe the same phenomenon at several levels without assuming that one level completely defines the others.
- CONCLUSION
A taxonomy of taxonomy provides a useful way to understand psychiatric classification.
Psychiatry contains multiple systems because classification serves multiple purposes. DSM-5-TR and ICD-11 provide established diagnostic and clinical classifications. HiTOP explores hierarchical dimensional organization of psychopathology. RDoC provides a research framework for studying behavioral and biological dimensions across conventional diagnostic boundaries.
These systems differ in what they classify, why they classify it, how their structures are organized, and what level of explanation they emphasize.
The central conclusion is therefore not that psychiatry requires only one final taxonomy. Rather, psychiatric knowledge can be conceptualized as a layered collection of taxonomies whose relationships themselves require classification.
In this sense, “taxonomy of taxonomy” becomes a legitimate psychiatric research problem: before psychiatry can determine how best to classify mental phenomena, it must also understand the kinds of classification it is using.
REFERENCES
[1] American Psychiatric Association. “About DSM-5-TR.” Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision, 2022.
[2] American Psychiatric Association. “What is the DSM?” American Psychiatric Association.
[3] World Health Organization. Clinical Descriptions and Diagnostic Requirements for ICD-11 Mental, Behavioural and Neurodevelopmental Disorders. WHO, 2024.
[4] National Institute of Mental Health. “About RDoC.” Research Domain Criteria.
[5] National Institute of Mental Health. “RDoC Matrix.” Research Domain Criteria.
[6] Cicero, D. C., Ruggero, C. J., Balling, C. E., et al. “State of the Science: The Hierarchical Taxonomy of Psychopathology (HiTOP).” Behavior Therapy, 55(6), 1114–1129, 2024. doi:10.1016/j.beth.2024.05.001.