Short Answer
A country profile rarely has room for the full linguistic geography of a state, so it compresses that reality into a short field: a list of languages, some status labels, and sometimes percentages. The compression is where confusion begins. Whether an indigenous or regional language appears at all — and how it appears — depends on which convention the profile follows: a census-style share of first-language speakers, a rank-ordered prevalence list, a legal list of official languages, or a country–language record in an atlas.
Why “indigenous” and “regional” are rarely field labels
The standard language field is legal and demographic rather than cultural. The World Factbook’s Languages entry, for example, is described as providing “a listing of languages spoken in each country” and specifying “any that are official national or regional languages” ([1](https://www.factbook.wiki/fields/402.html)). The vocabulary of the field is official, national, regional, first language, lingua franca — not a single indigenous/non-indigenous split.
That matters because two different questions are being asked. Indigenous describes a historical relationship between a community and a territory: a language spoken there by people whose presence long predates later settlement or conquest. Regional describes an administrative arrangement: a language recognised or used officially within part of a state, whether or not it is indigenous there. A language can be indigenous and regional, indigenous but unrecognised, regionally official but not indigenous (a former colonial language given provincial status), or neither. Most profile fields capture the second question well and the first only indirectly.
Four formats and what each one actually records
Percentage lists of first-language speakers
The most familiar format divides a population into shares. Where data are available, the Factbook field breaks languages down “according to the percent of the total population speaking each language as a first language” ([1](https://www.factbook.wiki/fields/402.html)). The denominator is the total population; the numerator is first-language speakers. It is not the share of the population that can speak the language, and it is not a count of ethnic identity.
Rank-ordered prevalence lists
Where no percentage data are available, the same field lists languages “in rank order based on prevalence, starting with the most-spoken language” ([1](https://www.factbook.wiki/fields/402.html)). This is an ordinal statement, not a measurement: it tells you that one language is more widely spoken than another, not by how much. Treating position one, two and three as if they were percentages is one of the most common misreadings of a country profile.
Legal official-language lists
A third format lists what the state recognises. The StreetComplete country metadata file, for instance, orders its entries from statewide official languages, through trade lingua francas, to written national majority languages and regionally recognised minorities; it records Andorra as [ca] (Catalan), Afghanistan as [fa, ps] (Dari and Pashto) and the United Arab Emirates as [ar] (Arabic) ([4](https://github.com/streetcomplete/countrymetadata/blob/master/data/officialLanguages.yml)). These entries describe status, not speaker numbers, and the file is a curated community dataset rather than an official government publication.
Country–language atlas records
UNESCO’s World Atlas of Languages structures its data “around two dimensions — Countries — Languages,” and its country profiles are built from information supplied by Focal Points appointed by UNESCO Member States, drawing on government institutions, national statistics offices, language commissions and higher-education bodies ([2](https://en.wal.unesco.org/countries)). Records of this kind can carry speaker numbers and vitality information, but comparability across countries depends heavily on what each national focal point supplied.
Reading percentages without misreading them
Afghanistan shows why a language field cannot simply be added up. In the 2017 estimate, the entry gives Afghan Persian or Dari at 77% (official, and functioning as the lingua franca), Pashto at 48% (official), Uzbek 11%, English 6%, Turkmen 3%, Urdu 3%, Pashayi 1%, Nuristani 1%, Arabic 1% and Balochi 1% ([1](https://www.factbook.wiki/fields/402.html)). Dari and Pashto alone already exceed 100%.
The source note explains that shares “sum to more than 100% because there is much bilingualism in the country and because respondents were allowed to select more than one language” ([1](https://www.factbook.wiki/fields/402.html)).
Two cautions follow. First, the percentages are selections, not a partition: a respondent could appear under both Dari and Pashto. Second, the entry states that the data represent the most widely spoken languages, so the list is not exhaustive — small languages may be absent without being unspoken.
The same distinction appears at global scale. In the 2025 edition’s world entry, the most-spoken languages (including second-language speakers) are led by English at 18.8% and Mandarin Chinese at 13.8%, while the most-spoken first languages are led by Mandarin at 12.3%, Spanish at 6%, and English and Arabic at 5.1% each (2022 estimates) ([3](https://thefactbook.org/fields/languages)). Switching the definition rearranges the ranking; the same is true inside any country profile.
What the status labels mean in practice
“Official” normally means the language is used by the state for law, administration or schooling. “National language” may be symbolic or may carry legal force, depending on the constitution. “Regional” signals official or recognised use within a subnational territory. “Lingua franca” describes a practical role in communication between groups, not a legal status.
Afghanistan compresses three tiers into one line: Dari and Pashto marked official, Dari additionally described as the lingua franca, and a note that Uzbek, Turkmen, Balochi, Pashayi, Nuristani and Pamiri are official in the areas where they are spoken ([1](https://www.factbook.wiki/fields/402.html)). That note is the closest the field comes to recording regional language status, and it is easy to miss when skimming a list of percentages.
Comparing the formats side by side
| Format | What it typically records | Useful for | What it cannot tell you |
|---|---|---|---|
| First-language percentage list | Share of total population speaking each language as a mother tongue, with some status tags | Order-of-magnitude comparisons within one country and reference year | Second-language speakers, ethnic identity, or shares that sum to 100% |
| Rank-ordered prevalence list | Relative order of languages where no shares are available | Rough hierarchy in data-poor settings | Any magnitude, trend or gap between ranks |
| Official-language list | Languages recognised by the state or a subnational territory | Legal and administrative status; comparing constitutions | How many people actually speak the language |
| Country–language atlas entry | Structured country and language records, often with speaker and vitality data | Deeper study of specific languages and communities | Uniform cross-country measurement, where national methods differ |
Country example: Afghanistan in two profiles
Put the two supplied sources side by side and the difference in intent is obvious. The official-language dataset reduces Afghanistan to two codes, [fa, ps] ([4](https://github.com/streetcomplete/countrymetadata/blob/master/data/officialLanguages.yml)). The Factbook entry gives ten languages with shares, two official tags, a lingua franca note and a regional-status sentence ([1](https://www.factbook.wiki/fields/402.html)). Neither tells you how many Uzbek speakers there are as a share of all Uzbek speakers, or how many people speak Uzbek as a second language. A reader who needs that must go to an atlas entry or the underlying national survey, not to a one-line country field.
Common interpretation errors
- Adding shares to 100%. Multilingual populations can produce totals well above 100%, as in Afghanistan.
- Reading rank as a share. A list ordered by prevalence gives no percentage, and the gaps between ranks are unknown.
- Confusing official with spoken. A language can be official while a minority speaks it, or widely spoken while unrecognised.
- Treating absence as zero. Fields often list only the most widely spoken languages, and a missing language is not evidence that it has no speakers.
- Assuming one reference year. Country entries are compiled from different censuses and estimates, so figures from two profiles may not be contemporaneous.
A short checklist for reading a language field
- Identify the format: percentages, rank order, legal status or atlas record.
- Note the observation year and whether the figure is a census, estimate or projection.
- Check the denominator: total population, adult population or first-language speakers only.
- Look for status tags — official, national, regional, lingua franca — and read any footnote.
- Check whether shares can overlap, and never force them to sum to 100%.
- Ask whether the list is explicitly partial before concluding a language is absent.
Limits of the evidence
These formats are not interchangeable, and the sources here do not resolve every question. Definitions of “regional language” vary by legal system, so a cross-country count of regionally official languages is not a standard statistical series. Census shares depend on self-identification and on whether respondents may choose more than one language. UNESCO’s atlas depends on national focal points, so coverage and detail differ between countries ([2](https://en.wal.unesco.org/countries)). And a field listing the “most widely spoken” languages is, by design, incomplete. Used with those limits in mind, country profiles remain a fast way to see which indigenous and regional languages a state recognises, which ones dominate demographically, and where the two answers diverge.
FAQ
Why do language percentages in a country profile sometimes add up to more than 100%?
Because respondents may report more than one language and because much of the population is bilingual. The Factbook's Afghanistan entry states this explicitly: Dari at 77% and Pashto at 48% alone exceed 100% under the 2017 estimate.
Does a missing language mean it is not spoken in that country?
No. Where data are available the field may list only the most widely spoken languages, and where data are absent it may give a truncated prevalence ranking. Absence is a gap in the field, not a measured zero.
What is the difference between an official-language list and a spoken-language list?
An official-language list records what the state recognises for law, administration or education; a spoken-language list records demographic shares. A language can be official while spoken by a minority, or widely spoken while unrecognised.
How does UNESCO's World Atlas of Languages collect its country data?
Its country profiles rely on information provided by Focal Points appointed by UNESCO Member States, drawn from government institutions, national statistics offices, language commissions and higher-education bodies, so detail and comparability vary by country.
Can I compare language shares across two different country profiles directly?
Only cautiously. Entries may come from different censuses, use different denominators and apply different status labels, so always check the observation year and definition before comparing.
Leave a Reply