SuperstitionsMap Data Methodology

Data Methodology: How the SuperstitionsMap Global Dataset Is Built

SuperstitionsMap is a comparative reference project about documented superstition, folk belief, oral tradition, legends, customary belief, omens, taboos, ritual traditions, proverbs, supernatural narratives and closely related folklore material around the world.

The dataset does not claim to measure how superstitious a population is, how strongly people currently believe a tradition, or the complete number of superstitions that exist in a country. No reliable worldwide census of superstitions exists.

The dataset measures something narrower and verifiable: the size of documented folklore and belief-related corpora that can be traced to identifiable sources and counted under a consistent publication methodology.

The current dataset covers 179 countries:

• 153 countries are ranked because a qualifying documented numerical corpus has been identified.
• 26 countries are Unranked because a defensible count has not yet met the evidence standard.

An Unranked status does not mean that a country has little folklore, few traditional beliefs or a weak oral tradition. It means only that the available evidence does not support a numerical value suitable for the comparative dataset.

The Core Interpretation Rule

The numbers on SuperstitionsMap should be read as documented corpus counts, not as literal totals of every superstition existing in a country.

A figure of 600 may represent 600 archived folktales. Another country’s figure may represent 600 documented taboos, belief records or legends. The precise unit is identified in the dataset and source register.

The available source base documents this many countable folklore or belief-related records under the stated scope.

It does not mean that the country has exactly that many superstitions.

The ranking is therefore a documentation index, not a ranking of populations by superstition, irrationality, religion or belief intensity.

What the Ranking Measures

Depending on the country and the surviving documentary tradition, a qualifying count may consist of:

• superstition entries;
• folk-belief records;
• omens;
• taboos;
• ritual-belief texts;
• legends;
• folktales;
• myths;
• supernatural narratives;
• culturally significant proverbs;
• oral-history narratives;
• traditional-custom records;
• archive records;
• field recordings;
• museum collection objects;
• clearly enumerated folklore texts;
• another defined traditional-culture corpus.

The original category used by the source is preserved whenever possible.

A collection of 1,000 proverbs is therefore described as a proverb and cultural-worldview corpus, not as “1,000 superstitions.” A collection of 1,000 legends remains a legend corpus. A museum collection remains a museum-object corpus.

What the Ranking Does Not Measure

• the percentage of people who believe in superstitions;
• current belief prevalence;
• religious belief;
• intelligence or education;
• irrationality;
• cultural sophistication;
• the complete amount of folklore existing in a society;
• the age of a folklore tradition;
• the popularity of individual beliefs;
• the number of people who know a tradition;
• the number of times a tradition has been recorded.

A high documented count may reflect a major historical collection project, strong archive preservation, extensive digitization, a broad dictionary, a detailed classification system or unusually intensive fieldwork.

A low documented count may simply reflect a narrower published source or limited digitization.

No country should therefore be described as “more superstitious” or “less superstitious” because of its position in the ranking.

No Population or Per-Capita Adjustment

The dataset is not normalized by population, land area, GDP, literacy rate or internet access.

A country with ten million inhabitants and a country with one million inhabitants may both have a documented corpus of 1,000 records. That does not imply equal prevalence of belief. It means that qualifying sources currently document corpora of similar numerical size.

Population normalization would introduce another unsupported assumption because the source corpora usually represent historical documentation rather than representative modern surveys.

Source Standard

Sources are prioritized according to provenance, transparency and reproducibility.

National academies, government agencies and official cultural institutions

These include ministries of culture, national folklore institutes, national academies of sciences, national heritage authorities, official language institutions, government-supported research projects and state cultural archives.

These sources are preferred when they clearly state the size and scope of a corpus.

Universities, research institutes, libraries, archives and museums

Institutional repositories and specialist collections are major sources for the dataset.

• national folklore archives;
• university folklore collections;
• museum catalogues;
• university theses and dissertations;
• research-centre databases;
• national and university library catalogues;
• institutional digitization projects.

Peer-reviewed and established academic publications

Academic journals, scholarly books, critical editions and research monographs are accepted when they identify the corpus clearly enough to determine what is being counted.

Specialist cultural and heritage organizations

Recognized cultural foundations, folklore organizations and heritage projects may be used when their corpus is clearly documented and the count is reproducible.

Historical primary collections

Older folklore collections remain valuable when their provenance is known, the collecting context is identifiable, the number of records is explicit or reproducible, and modern scholarly or bibliographic sources can verify the work.

A nineteenth- or early twentieth-century source is not rejected merely because of age. Historical folklore research is often the only surviving record of oral traditions that have since changed or disappeared.

Multi-source documented minimums

For some countries no single national corpus provides a useful total. A lower-bound corpus may be constructed from multiple academic or institutional sources only when:

• the individual traditions are identifiable;
• the sources are relevant to the country;
• obvious duplicates are excluded;
• the resulting figure is explicitly described as a minimum;
• the number is not presented as a complete national total.

Sources That Are Not Accepted as Numerical Evidence

• anonymous webpages;
• generic travel blogs;
• unsourced lists;
• social-media posts;
• forum claims;
• commercial listicles;
• AI-generated summaries;
• search-engine snippets;
• automatically generated statistics without source provenance.

Search engines may help locate an underlying publication, archive or catalogue, but the search result itself is not treated as evidence.

Research and Verification Workflow

1. Country-specific source discovery

Research uses combinations of the country’s name with concepts such as superstition, folk belief, popular belief, folklore, folk tradition, oral tradition, legend, folktale, myth, omen, taboo, ritual, proverb, custom, ethnography and supernatural belief.

Local-language equivalents are also used where they improve source discovery.

2. Institutional and academic sources are identified

Candidate sources are evaluated according to institution, authorship, publication history, research context, collection scope and bibliographic traceability.

3. The underlying source is checked

The numerical statement must refer to an identifiable corpus. The presence of a number somewhere on a page is not enough.

4. The counted unit is identified

The research must establish whether the number represents separate entries, tales, legends, belief statements, taboos, proverbs, recordings, manuscripts, archive units, objects or another defined unit.

5. Geographic scope is identified

The source may be national, regional, provincial, island-specific, ethnic, language-specific, thematic or collection-specific. The scope is retained in the description.

6. Possible duplication is evaluated

A later publication may reproduce an earlier corpus. Multiple databases may also contain the same original records. Such material is not intentionally added together unless the collections can be shown to be distinct.

7. The conservative reproducible number is preferred

When several interpretations are possible, the figure requiring the fewest assumptions is selected. A smaller directly verified figure is preferable to a larger inferred figure.

8. Provenance is recorded

For ranked data, the dataset records as much of the following as applicable:

• published count;
• count type;
• source title;
• source institution;
• source URL;
• explanation of scope;
• methodological limitation;
• reason the corpus is suitable for the ranking.

9. Ranking requires numerical qualification

If the count does not meet the standard, the country remains Unranked. No number is invented merely to complete the ranking.

Numbers Deliberately Rejected

Page counts

A 300-page folklore book does not equal 300 folklore traditions. Page count describes the physical size of a publication, not its number of records.

Participant counts

A study involving 600 participants does not equal 600 beliefs. Respondents may report multiple traditions, the same tradition, or none.

Interview counts

One interview can contain dozens of traditions. Ten interviews do not equal ten beliefs.

Word counts

The number of words in a corpus says nothing reliable about the number of separate folklore traditions.

Archive size when genres are mixed

An archive may contain songs, photographs, administrative papers, jokes, children’s games, tales, customs, beliefs and sound recordings. The entire archive cannot automatically be converted into a superstition or folklore-item count.

Publication volume counts

A six-volume collection does not mean six traditions. The actual number of records must be known.

Unverified estimates

A culturally plausible estimate is not accepted solely because it appears reasonable.

General statements such as “thousands”

If a source says only that “thousands of traditions” exist without giving a usable lower bound or corpus definition, the statement is normally unsuitable for ranking.

Duplicate publications

The same tale reproduced in three books is not three different traditions.

Exact, Approximate and Lower-Bound Numbers

Exact-style count: 600

The source states a specific total or the corpus can be reproducibly counted to that figure. It still represents the stated corpus, not necessarily the complete folklore of the country.

Documented lower bound: 600+

At least 600 records are supported, while available evidence shows that the corpus extends beyond that minimum. The plus sign must remain visible.

Approximate corpus: ≈600

The source describes the corpus approximately or the underlying database supports a close but not perfectly fixed figure. The approximation sign must not be silently removed.

National Corpora, Regional Corpora and Documented Minimums

A national corpus is preferred, but folklore documentation rarely follows modern political boundaries perfectly.

Important collections may cover a province, particular island, historical region, ethnic group, language community, city, research theme or one class of supernatural belief.

Such corpora can still provide meaningful documented evidence. When they are used, they are treated as scoped documented corpora or documented minimums. They are not represented as complete national totals.

For example, a regional collection of 500 legends establishes that a substantial documented corpus exists for the country. It does not establish that the whole country possesses exactly 500 legends.

Historical Borders and Modern Countries

Historical folklore collections sometimes predate modern national boundaries.

A source is associated with a present-day country only when the geographic, linguistic or cultural connection is sufficiently clear.

Where a collection covers only a historical region now divided between countries, the scope is stated rather than silently treating the entire historical corpus as belonging to one modern state.

Ethnic and Language-Specific Corpora

Traditional knowledge is often recorded by ethnic or linguistic community rather than nationality.

Such corpora may be included when the community is clearly associated with the country, the source location is identifiable, the scope is disclosed and the corpus is not falsely labeled as a complete national total.

Why Proverbs Appear in the Dataset

Proverbs are not automatically treated as superstitions.

Historical proverb collections may preserve traditional concepts of luck, warnings, agricultural expectations, supernatural assumptions, social prohibitions, customary beliefs, ideas about fate and culturally inherited explanations of events.

A proverb corpus may therefore serve as a traditional cultural-worldview corpus when supported by scholarly context. Such data are labeled as proverb or cultural-worldview material rather than being renamed as individual superstitions.

Generic modern quotation collections do not qualify.

Legends, Myths and Folktales

Many historical folklore archives do not divide oral tradition into a modern category called “superstition.” Beliefs about spirits, omens, witches, sacred places, dangerous beings, miraculous events, taboo violations and supernatural consequences may survive primarily in legends and traditional narratives.

The dataset retains the original genre and describes the number as a legend, tale or narrative corpus.

Archive Units

Some institutions count archival units rather than individual narratives. An archival unit may contain one or more documents.

When such a figure is used, the dataset explicitly describes it as archival units. It is not converted into a claim about an equal number of unique superstitions.

Museum Objects

A museum collection can preserve material evidence relating to beliefs and customs. When a museum provides a defined collection count, that count may be used as a collection measure, but the objects remain museum objects rather than being relabeled as superstition entries.

Audio and Field Recordings

Historical sound archives are important sources of oral tradition.

A recording count may be used when the archive is clearly folklore-related, the collection scope is known and the number refers to a defined body of recordings.

A cassette or recording may contain more than one narrative, so recording counts are retained as recording counts.

Mixed-Genre Collections

Many folklore collections combine beliefs, legends, tales, jokes, ritual customs, songs, proverbs and oral history.

Mixed corpora can qualify when they remain clearly within traditional folklore or cultural-worldview documentation. Their count descriptions state that broader scope.

Multi-Source Research Corpora

A small group of countries use documented minimums assembled from several scholarly or institutional sources rather than one enumerated national database.

These are intentionally conservative. The process requires:

1. identifying individual documented traditions;
2. tracing each tradition to a credible source;
3. excluding obvious repetitions;
4. counting only the surviving distinct documented items;
5. publishing the result as a minimum rather than a national total.

A documented minimum of 60+ means that at least 60 source-supported items have been established. It does not estimate the unknown remainder.

Why Larger Numbers Are Not Automatically Preferred

A larger figure is not automatically more authoritative.

If one interpretation suggests 290 records but a direct academic source explicitly verifies 133, the direct 133-record corpus is preferable until the larger figure can be independently confirmed.

Better provenance before bigger numbers.

Ranking Rules

Only countries with a qualifying numerical value receive a rank.

Ranked countries are ordered by the numerical value underlying the published documented count. Approximation and lower-bound symbols remain visible.

Countries without a qualifying value are not placed artificially beneath the lowest ranked country. They are marked Unranked.

This prevents a lack of numerical evidence from being interpreted as a very small folklore tradition.

Ties and Near-Ties

The ranking should not be interpreted as a high-precision statistical league table.

A country with 500 documented records and a country with 496 documented records cannot meaningfully be described as culturally different because of a four-record gap.

The corpora were created at different times, collected for different purposes, classified differently and preserved by different institutions.

The ranking is most useful for understanding the scale of documented source corpora.

Unranked Status Labels

Verified count unavailable

Relevant traditional material may exist, but no qualifying enumerated corpus has been verified.

Corpus identified · Count unavailable

A credible folklore corpus is known, but the available source does not state a usable item total.

National count unavailable

Large regional collections may exist, but they are not treated as national totals.

Count awaiting institutional verification

A numerical corpus has been identified, but confirmation from the preferred institutional or primary source is still required.

Count awaiting primary verification

The underlying project is credible, but the numerical total currently depends on a secondary description rather than the preferred primary record.

Count requires verification

A reported number may be legitimate, but the evidence chain does not yet justify ranking.

Documented subset only

A small verified corpus exists, but it is too narrowly selected to function responsibly as the country’s ranking value.

Curated subset available

An official museum or heritage body may present a small curated selection. The number is real, but the selection should not be mistaken for the national corpus.

Count not used for ranking

A publication contains a numerical corpus, but methodological weaknesses make it unsuitable for the comparative ranking.

Not enough country-specific data

Relevant archives may exist, but their material cannot be separated reliably into a country-specific corpus.

Data Fields

• Rank: Position among countries with qualifying documented counts.
• Documented count: Published numerical value, including + or ≈ where required.
• Count type: Description of exactly what the number measures.
• Source title: Publication, collection, database or project supporting the value.
• Source institution: University, archive, museum, government agency, publisher or research body responsible for the source where identifiable.
• Source note: Explanation of what the source says, what the number represents, geographic scope, genre and limitations.
• Source URL: Link to the evidence record, institutional catalogue, publication or accessible source page.
• Unranked status: Reason a country does not currently receive a numerical rank.

Country-by-Country Ranked Data Register

Every value below describes the corpus actually used in the ranking. It does not claim to represent the complete number of superstitions existing in that country.

Countries Currently Unranked

These countries remain part of the global map and directory but do not receive numerical positions until the evidence standard is satisfied.

Corrections and Revisions

Folklore databases change. Archives are digitized. New catalogues become available. Historical publications become searchable. Existing collections may also be reclassified.

A country value may therefore change when stronger evidence becomes available.

A revision should favor:

1. stronger provenance;
2. clearer corpus scope;
3. a more reproducible count;
4. a better institutional source;
5. a more conservative interpretation.

A revised number can be lower than the previous number. That is not a methodological failure. It can represent an improvement in evidence quality.

If an earlier interpretation suggested 290+ records but a directly verifiable academic source establishes 133, the documented 133-record corpus should be preferred until the larger total can be confirmed independently.

Broken or Disappearing Sources

Web pages can disappear even when the underlying book, archive or research project remains valid.

When a source URL disappears, the evidence should be re-established through one or more of:

• the original institution;
• an institutional repository;
• a library catalogue;
• a DOI record;
• a national bibliography;
• a preserved publication;
• an authoritative archive.

If the numerical basis can no longer be defended, retaining an unsupported rank is less appropriate than revising the figure or moving the country to Unranked.

Corrections Do Not Require a Larger Number

The objective is not to maximize every country’s count. The objective is to make every published count defensible.

A correction may increase a number, decrease a number, add an approximation symbol, convert an exact number into a lower bound, change the description of what is counted, replace a secondary source with a primary source or remove a country from the ranking.

Limits of Cross-Country Comparison

There is no universally standardized academic unit called a “superstition record.”

Different folklore traditions have been documented using different systems. One archive may catalogue every legend separately. Another may catalogue one manuscript containing dozens of legends. A dictionary may count individual headwords. A field researcher may count complete narratives. A museum may count objects. A sound archive may count recordings.

These numbers cannot be treated as if they came from one standardized global survey.

SuperstitionsMap therefore makes the unit visible instead of hiding the difference.

The dataset is best interpreted as a map of documented folklore and traditional-belief corpus scale.

Why Some Countries May Remain Unranked

A country can possess a rich oral tradition while remaining Unranked for years.

• major collections have never been digitized;
• catalogues exist only in print;
• archives do not publish item totals;
• historical research combines several modern countries;
• local-language material is difficult to access;
• a collection mixes many genres without item-level breakdown;
• institutional websites provide descriptions but no count;
• a reported number exists only in a weak secondary source;
• available collections are too narrowly curated to support a comparative value.

In such cases, not publishing a number is the more informative result.

Methodological Principle

Source first, number second.

No country requires a number merely for the sake of making the dataset appear complete.

Every ranked value should be explainable through four questions:

1. What exactly was counted?
2. Which source supports the count?
3. What geographic or thematic area does the corpus cover?
4. What limitations apply to the number?

If those questions cannot be answered adequately, the country remains Unranked.

SuperstitionsMap functions as a living documentation index. Its authority depends not on making every cell look complete, but on keeping the difference between documented evidence, approximation, lower bounds and unavailable data visible.

A transparent missing value is stronger evidence practice than a precise-looking number with no defensible source.