L-MAS — LINGACON Multilingualism Assessment Standard
How many languages does the average person actually speak?
A standardized framework for measuring multilingualism at population level
We know how many languages exist on Earth (over 7,000). We know which ones are endangered. We know their grammar, their phonology, their history. But comparable data on the most basic fact about human linguistic capacity — how many languages a person actually speaks — has been missing.
Dozens of surveys, censuses, and databases collect language data — but they use different questions, different thresholds, and different definitions. A person counted as "bilingual" in one survey may not qualify in another. Results from different countries cannot be compared. Entire regions — the Arab states, Latin America, most of Asia — have no publicly available standardized multilingualism data at all. L-MAS was developed to close this gap.
The LINGACON Multilingualism Assessment Standard (L-MAS) is a five-level framework for measuring multilingualism at population level. It scales from a single survey question (Level 1, zero cost) to a full national longitudinal tracker (Level 5). Every level produces the same core metric — ALPS (Average Languages a Person Speaks) — ensuring that results are comparable across countries, time periods, and survey programs.
L-MAS is free, open, and designed for immediate adoption. The Level 1 question can be added to any existing survey today.
Each level builds on the previous one. Organizations choose the level that matches their resources and objectives.
| Level | Name | Items | Primary Output | Cost |
|---|---|---|---|---|
| Level 1 | The One Question | 1 question | National ALPS estimate | Zero |
| Level 2 | The Module | 4 questions | ALPS + proficiency depth | Minimal |
| Level 3 | The Full Battery | 8–12 questions | ALPS + connectivity data | Moderate |
| Level 4 | National Audit | Standalone survey | Gold-standard ALPS + CI | Variable |
| Level 5 | Longitudinal Tracker | Level 3 repeated | ΔALPS over time | Sustained |
L-MAS is part of the LINGACON research program on global linguistic connectivity. The following papers document the framework and its applications.
Hovhannisyan, A. (2026). The LINGACON Multilingualism Assessment Standard. LINGACON Project.
A global estimate of average languages a person speaks across 235 countries and territories.
How disconnected is humanity? Measuring pairwise linguistic connectivity across 8.3 billion people.
Multilingualism and linguistic connectivity, measured to a common standard
We conduct L-MAS assessments for organizations, cities, and countries — delivering two standardized metrics: ALPS (how many languages your people speak, on average) and GLCI (how linguistically connected your population is).
For multinational companies operating across 3+ countries. Measure your workforce's multilingualism and internal linguistic connectivity. Identify communication gaps and language training priorities.
For municipalities, states, provinces, and metropolitan areas. Understand the linguistic landscape of your population. Plan integration services, education policy, and public communication in the right languages.
For governments and national institutions. A full L-MAS Level 3 or Level 4 assessment producing publication-quality ALPS data with confidence intervals, regional breakdown, and policy recommendations.
See what an L-MAS assessment delivers — fictional company, real methodology.
Select your profile below. We'll ask the right questions and respond with a detailed proposal.
An intuitive assumption is that larger countries are harder and more expensive to survey. The data reveals the opposite: language diversity, not population size, determines the complexity and cost of multilingualism measurement.
Japan has about 124 million people and the vast majority of the population is monolingual. A representative multilingualism survey requires a standard national sample of 2,000–3,000 respondents.
Rwanda has 14 million people but extreme linguistic diversity — Kinyarwanda, French, English, Swahili, and dozens of local languages in varying combinations. A representative sample must be stratified across linguistic regions, urban/rural divides, and education levels. The sample size may need to reach 8,000–12,000 respondents, the interviewer training becomes more complex, and the data cleaning more labor-intensive.
The result: measuring multilingualism in Rwanda — a country one-ninth the population of Japan — can cost significantly more per survey point than measuring it in Japan.
This principle applies at every scale. A linguistically diverse city (Toronto, Dubai, Singapore) is harder to assess than a homogeneous one. A multinational company operating across 15 countries needs more sophisticated measurement than one operating in 3.
Understanding this relationship is critical for budgeting, planning, and interpreting L-MAS assessments. We provide detailed cost estimates tailored to your specific linguistic landscape.
For research inquiries, partnership proposals, media requests, or any other questions.