Research

Language data is abundant and hard to compare. One census asks which languages you speak at home; a survey next door asks which you "know"; a third records only your mother tongue and your best foreign language. A person counted as bilingual under one instrument may not qualify under another, and it is often unclear whether two national figures describe the same thing.

The LINGACON research programme addresses this in three steps: a measurement standard, indices built under it for every country and territory, and open publication of the results with everything needed to check them. Each step is documented in its own paper, released with its data.

How we work

One definition, everywhere

Every estimate is produced under one explicit threshold, conversational competence at CEFR A2 and above, and one set of counting rules. Sources that use other thresholds are harmonized to ours, and each adjustment is documented.

Every number traceable

Each national figure is assembled from that country's best available evidence and checked against independent sources. Where sources disagree, we report the disagreement instead of averaging it away, and we state which way the remaining uncertainty is likely to point.

Open by default

Papers are released with their data, code, assumption registers and sensitivity analyses, so that a reader can check every figure.

Publications

Preprints on SocArXiv under CC BY 4.0, with the data and code on the Open Science Framework.

Standard · l-mas.com

L-MAS: The LINGACON Multilingualism Assessment Standard

L-MAS is a five-level framework for measuring multilingualism at population level, from a single survey question (Level 1) through a four-question module and a full battery to a national audit and a longitudinal tracker. Every level produces the same core metric, ALPS, so that a country measuring at Level 1 and one measuring at Level 4 produce comparable numbers.

The standard is open and aligned with the CEFR, and the Level 1 question can be added to a survey already in the field. The five levels and the full Level 1 question are described on the standard's own site.

Read the standard at l-mas.com

L-MAS assessments of a company, a city or a country can be requested; see Partnerships.

Preprint · version 2 in moderation

The LINGACON Multilingualism Assessment Standard (L-MAS): Definitions, Thresholds, and a Review of Survey Practice

Hovhannisyan, A. (2026b). The LINGACON Multilingualism Assessment Standard (L-MAS): Definitions, thresholds, and a review of survey practice. SocArXiv. Preprint page (latest version): https://osf.io/preprints/socarxiv/n96g3. Version 1: https://doi.org/10.31235/osf.io/n96g3_v1

The definitions and thresholds behind the standard, and a review of how existing surveys ask about languages. Version 2, which corrects the reading of the Afrobarometer Round 9 and Round 10 items on other spoken languages, is awaiting moderation; version 1 remains available.

Preprint · SocArXiv

The ALPS Index: A Global Estimate of the Average Number of Languages a Person Speaks

Hovhannisyan, A. (2026a). The ALPS Index: A global estimate of the average number of languages a person speaks. SocArXiv. https://doi.org/10.31235/osf.io/3mtsx_v1

A population-weighted estimate of ALPS (Average Language a Person Speaks) for 237 countries and territories, with 934 languages indexed: 1.57 languages per person (1.568), within a band of 1.42–1.70. The paper documents the sources behind every national figure, the harmonization that reconciles different instruments, the audit trail, and the uncertainty around the global estimate, which is reported as a band.

Preprint · SocArXiv

GLCI (Global Linguistic Connectivity Index): How Connected Is the World?

Hovhannisyan, A. (2026c). GLCI (Global Linguistic Connectivity Index): How connected is the world? SocArXiv. https://doi.org/10.31235/osf.io/g3vxj_v1

The probability that a randomly chosen listener understands a randomly chosen speaker, computed across the world's population and decomposed into pairs within countries and across borders, by region, and by the contribution of individual bridge languages. Worldwide it is 11.3%; across borders, 7.5%. English carries 39% of this connectivity.

Data and code · Open Science Framework

Data, code and supplementary materials

https://osf.io/7bshn

Supplementary materials, row-level data, the pair table and the analysis code for the ALPS and GLCI papers. Licences are stated for each component.

In preparation

Paper 3: in preparation

On what linguistic connectivity means for societies.

Planned studies

The following are on the research agenda and will proceed as resources allow. They are listed here because researchers working on the same ground are welcome to get in touch.

Planned

ALPS and GLCI for the states of the United States

National averages conceal large internal variation. Measuring each US state separately would show how a country's linguistic connectivity is distributed inside its own borders, and how much a single national figure says about a federation of that size.

Planned

India: diversity and connectivity

India is among the most linguistically diverse places on Earth. Measuring every state separately would show how that diversity relates to connectivity inside the country and with the rest of the world.

Planned

China: size and reach

Mandarin has more first-language speakers than any other language, yet its reach across China's borders is limited. How the size of a language relates to its reach across borders is a question worth studying in its own right.

Interested in supporting or collaborating on any of these? Write to research@lingacon.com, or see Contribute.