Lexical Half-life

This article relies on measurements that may have decayed since publication. The half-lives reported herein are themselves subject to lexical decay. (January 2026)

Lexical Half-life
Also known as	Word decay constant, Semantic radioactivity, Definitional instability
Field	Linguistics, Drift theory, Information theory
First formalized	2019
Formalized by	Dr. Ingrid Solheim
Symbol	t½ or λlex
Unit	Years (lexical)
Related concepts	Semantic drift, Temporal debt, Radioactive decay

Lexical half-life (t_½) is a measure of the time required for a word's original meaning to decay to half of its initial definitional stability. First formalized by Norwegian computational linguist Dr. Ingrid Solheim at the University of Oslo in 2019, the concept applies principles from nuclear physics to the study of semantic drift, proposing that words undergo a form of "definitional radioactivity" wherein their meanings spontaneously decay according to predictable statistical laws.^[1]

Unlike traditional approaches to semantic change, which focus on the social and cultural factors driving meaning evolution, lexical half-life theory treats semantic instability as an intrinsic property of words themselves—a linguistic analog to radioactive isotopes, where some words are inherently more stable than others regardless of usage context.^[2]

Origins[edit]

The concept of lexical half-life emerged from a 2017 collaboration between linguists and physicists at the Oslo Centre for Interdisciplinary Studies. Researchers attempting to model semantic drift rates noticed that the decay curves of certain word meanings closely resembled exponential decay functions familiar from radiochemistry.^[3]

Dr. Solheim, then a postdoctoral researcher, recognized that if meanings decayed according to consistent mathematical laws, individual words might possess characteristic "decay constants" analogous to those of radioactive isotopes. Her 2019 paper, "On the Radioactivity of Language," demonstrated that historical semantic change data for over 3,000 English words could be fitted to exponential decay models with statistically significant accuracy.^[4]

"We do not claim that words literally emit particles or radiation. But the mathematics are strikingly parallel. Meanings decay. They do so at characteristic rates. And like radioactive decay, the process appears to be fundamentally probabilistic—we cannot predict which specific meaning component will decay next, only the statistical likelihood across a population of usages."
— Dr. Ingrid Solheim, 2019

Theoretical framework[edit]

Decay mechanics

According to lexical half-life theory, every word possesses a quantity Solheim termed definitional mass (M_d)—the totality of meaning components, connotations, and semantic boundaries that constitute the word's full definition at a given moment. This definitional mass undergoes continuous decay, with meaning components spontaneously "detaching" from the core definition and either dissipating entirely or reattaching to other words.^[5]

The decay process follows the standard exponential decay equation adapted for linguistic contexts:

M(t) = M₀e^−λt

Where M(t) is the definitional mass at time t, M₀ is the initial definitional mass, λ is the decay constant (specific to each word), and e is Euler's number. The half-life is then calculated as:

t_½ = ln(2) / λ

Stability classes

Solheim's research identified five primary stability classes for English words, based on their characteristic half-lives:

Class	Half-life range	Examples	Characteristics
Ultra-stable	> 500 years	water, mother, three, eye	Core vocabulary, concrete referents, universal human experience
Stable	100–500 years	justice, love, honor, truth	Abstract but deeply embedded concepts
Moderate	25–100 years	success, freedom, professional	Culturally significant terms subject to generational reinterpretation
Unstable	5–25 years	literally, problematic, curate	Terms undergoing active semantic negotiation
Volatile	< 5 years	based, mid, slay	Slang, internet terminology, rapidly evolving jargon[6]

Measurement methodology[edit]

Calculating lexical half-life requires tracking changes in definitional mass over time. Solheim developed the Definitional Mass Spectrometry (DMS) method, which quantifies meaning through several metrics:^[7]

• Synonym overlap coefficient: The degree to which a word's synonyms remain constant over time
• Contextual distribution shift: Changes in the syntactic and semantic contexts where the word appears
• Definition entropy: The variance in definitions provided by native speakers in controlled elicitation studies
• Collocational stability: Persistence of characteristic word combinations

These metrics are combined into a composite Definitional Integrity Index (DII), which is then tracked longitudinally. When the DII falls to 50% of its baseline value, one half-life has elapsed.

For historical analysis, researchers employ corpus linguistics techniques, comparing word usage patterns across dated text corpora spanning centuries. For contemporary measurements, the Oslo Lexical Decay Observatory conducts annual surveys tracking real-time meaning changes in over 10,000 English words.^[8]

Observed half-lives[edit]

The following table presents half-life measurements for selected words, based on data from the Oslo Lexical Decay Observatory's 2024 report:^[9]

Word	Measured half-life	Decay status	Notes
literally	8.3 years	Active decay	Original meaning ~40% remaining; intensifier usage now dominant
decimate	67 years	Post-decay stable	Original "reduce by one-tenth" effectively extinct; new meaning stabilized
awesome	42 years	Post-decay stable	Sacred/fearful connotations fully decayed; casual positive meaning dominant
algorithm	12.7 years	Active decay	Technical definition expanding into general "opaque system" meaning
trauma	18.4 years	Active decay	Clinical specificity eroding; expanding to encompass minor distress
sustainable	15.2 years	Active decay	Environmental meaning diffusing into general "good practice" sense

Factors affecting decay rate[edit]

While lexical half-life is treated as an intrinsic property, research has identified several factors that can accelerate or retard decay:

• Usage frequency paradox: Highly frequent words tend toward longer half-lives, but sudden spikes in usage (as when a word "goes viral") can trigger accelerated decay^[10]
• Referent stability: Words with stable, concrete referents decay more slowly than those referencing shifting social or technological phenomena
• Institutional anchoring: Words with meanings fixed by legal, scientific, or religious institutions resist decay longer than vernacular terms
• Cross-linguistic borrowing: Words borrowed between languages may experience "decay reset" as they acquire new definitional mass in the receiving language
• Temporal debt accumulation: Some researchers have proposed that collective temporal debt may accelerate lexical decay across entire vocabularies, though this hypothesis remains speculative^[11]

The relationship between lexical half-life and semantic drift remains an area of active investigation. Solheim has proposed that drift represents the "directional component" of decay—decay determines how quickly meaning changes, while drift determines in which direction it moves.^[12]

Applications[edit]

Lexical half-life theory has found applications in several domains:

• Legal interpretation: Courts have begun considering half-life data when interpreting historical statutes, recognizing that key terms may have undergone significant decay since drafting^[13]
• Translation and localization: Translators use half-life estimates to identify terms at high risk of meaning shift during long-term projects
• Brand naming: Marketing firms consult decay predictions to select product names with stable semantic profiles
• Artificial intelligence: Language model developers use half-life data to weight training corpora, reducing influence of texts containing decayed meanings^[14]
• Lexicographic dating: Half-life analysis provides new methods for dating undated historical texts based on semantic decay signatures

A particularly intriguing application has emerged in the study of what researchers term ghost vocabulary—words whose meanings have fully decayed but which persist in usage as semantic shells, their definitional mass approaching zero while their forms remain in circulation.^[15]

Criticism[edit]

Lexical half-life theory has attracted significant criticism from linguists who reject its core premises:

• Critics argue that the analogy to radioactive decay is superficial and misleading, as physical decay involves actual particle emission while semantic change is a social phenomenon with no comparable mechanism^[16]
• The concept of "intrinsic" decay rates has been challenged as reifying what are actually socially contingent processes of meaning negotiation
• Some researchers have failed to replicate Solheim's half-life measurements, obtaining widely varying results depending on corpus selection and measurement methodology
• Philosophers of language have questioned whether "definitional mass" is a coherent concept, arguing that meaning cannot be meaningfully quantified^[17]

Solheim has responded to critics by emphasizing that the theory is explicitly metaphorical and does not claim that words literally undergo radioactive decay. She maintains that the mathematical parallels are empirically robust regardless of underlying mechanism: "The equations work. The predictions hold. Whether we call it decay or drift or dance, the patterns are real."^[18]

See also[edit]

• Semantic drift
• Temporal debt
• Mnemonic commons
• Consciousness archaeology
• Semantic change
• Radioactive decay
• Ghost vocabulary
• Chronolinguistics
• Oslo Lexical Decay Observatory

References[edit]

1. ^ Solheim, I. (2019). "On the Radioactivity of Language: Toward a Theory of Lexical Half-life". Journal of Quantitative Linguistics. 26 (3): 201–234.
2. ^ Solheim, I. (2020). The Decay of Words: Lexical Half-life and the Mathematics of Meaning. Oslo: Scandinavian University Press.
3. ^ Hansen, K.; Solheim, I. (2017). "Exponential Models of Semantic Change: A Preliminary Investigation". Oslo Working Papers in Linguistics. 42: 1–28.
4. ^ Solheim, I. (2019). "Fitting Decay Curves to Historical Semantic Data". Computational Linguistics. 45 (2): 289–312.
5. ^ Solheim, I. (2021). "Definitional Mass and Semantic Decay Mechanics". Theoretical Linguistics. 47 (1): 45–78.
6. ^ Solheim, I.; Tanaka, Y. (2022). "A Taxonomy of Lexical Stability Classes". Language Dynamics and Change. 12 (2): 156–189.
7. ^ Oslo Lexical Decay Observatory (2020). "Definitional Mass Spectrometry: Technical Protocols". OLDO Technical Reports. 3: 1–67.
8. ^ Oslo Lexical Decay Observatory (2021). "Methodology for Real-Time Half-life Measurement". OLDO Technical Reports. 7: 1–45.
9. ^ Oslo Lexical Decay Observatory (2024). Annual Report on English Lexical Decay. Oslo: OLDO Publications.
10. ^ Chen, S.; Solheim, I. (2023). "The Viral Decay Effect: Usage Spikes and Accelerated Half-life Reduction". Internet Linguistics Quarterly. 8 (1): 34–56.
11. ^ Voss, H.; Solheim, I. (2022). "Temporal Debt and Vocabulary-wide Decay Acceleration". Journal of Chronopsychology. 47 (2): 178–201.
12. ^ Solheim, I.; Fontaine, M. (2023). "Decay and Drift: Toward an Integrated Framework". Cognitive Semantics. 31 (1): 89–123.
13. ^ Morrison, J. (2024). "Lexical Half-life in Statutory Interpretation: A Framework". Yale Law Journal. 133 (4): 901–956.
14. ^ Zhang, W.; et al. (2025). "Temporal Weighting in Language Model Training Using Decay Constants". Proceedings of ACL 2025: 456–478.
15. ^ Solheim, I. (2025). "Ghost Words: When Definitional Mass Approaches Zero". Language. 101 (1): 45–78.
16. ^ Williams, P. (2021). "Against Lexical Half-life: A Critique of Pseudo-Physical Linguistics". Philosophy of Language Quarterly. 64 (3): 267–289.
17. ^ Brandt, E. (2022). "Can Meaning Be Measured? Challenges to Definitional Mass Theory". Theoretical Linguistics. 48 (3): 312–334.
18. ^ Solheim, I. (2023). "Response to Critics: The Empirical Case for Lexical Half-life". Journal of Quantitative Linguistics. 30 (1): 78–99.