Language Representation in Alma CZ Bibliographic Records Using ISO 639-2 and ISO 639-3 Codes

Following the February 2026 Alma release, support for ISO 639-3 language codes was introduced as part of a combined controlled vocabulary for languages.

As a result, Alma Community Zone (CZ) bibliographic records may include multiple language codes representing the same language, using both:

ISO 639-2 (MARC standard)
ISO 639-3 (expanded language set)

This article explains the expected behavior and its impact.

Background

The February 2026 enhancement introduced:

A merged language vocabulary combining:
- ISO 639-2
- ISO 639-3
Usage in:
- Language fields within bibliographic records
- Metadata Editor
- Discovery indexing and facets

This enables more granular and comprehensive language representation in Alma.

Behavior in CZ Bibliographic Records

In Alma CZ bibliographic records, it is possible to encounter:

Multiple language entries representing the same language
Different ISO standards used in parallel

Example

A CZ bibliographic record may include:

Language: fre
Language: fra

or:

Language: ara
Language: arb

Explanation

CZ records aggregate metadata from multiple sources and enrichment processes
These sources may use:
- ISO 639-2 codes (e.g., fre, ara)
- ISO 639-3 codes (e.g., fra, arb)

Alma preserves all values present in the CZ bibliographic record.

Expected Behavior

This behavior is expected and by design.

Language fields in CZ bibliographic records may contain multiple values
Alma does not deduplicate or merge language codes
All language values are:
- indexed
- searchable
- exposed in discovery

Even if the values represent the same language at different levels of granularity

Impact on Discovery

In Primo VE:

All language values from CZ bibliographic records are indexed
Language facets may display:
- multiple entries for the same language
- both general and specific forms

Example

Code	Display
ara	Arabic
arb	Standard Arabic

Both may appear simultaneously in the facet.

Why Deduplication Is Not Performed

Alma does not merge or deduplicate ISO language codes in CZ bibliographic records because:

CZ is designed to preserve source metadata integrity
ISO 639-2 and ISO 639-3 represent different levels of linguistic granularity
Automatic merging could:
- remove valid distinctions
- introduce inconsistencies
- impact search and indexing behavior

Recommendations

Institutions working with CZ bibliographic records may choose to:

Accept the default behavior

Retain both ISO standards
Benefit from richer language representation

Apply local normalization

Standardize language codes (e.g., ISO 639-2 only)
Map or remove duplicate representations
Implement via:
- normalization rules
- local workflows

Adjust discovery behavior

Use Primo VE normalization rules to:
- merge or collapse language facet values
- reduce duplication in display

Important Notes

CZ bibliographic records may include mixed language standards
No automatic alignment is performed between:
- different ISO standards
- different language fields
This behavior applies specifically to CZ-managed metadata

Conclusion

In Alma Community Zone (CZ) bibliographic records, the presence of both:

ISO 639-2
ISO 639-3

for the same language is:

✔️ Expected
✔️ A result of expanded language support introduced in February 2026
✔️ Not automatically normalized or deduplicated

Institutions requiring consistency should implement local normalization or discovery adjustments.

Internal Note

This article addresses a documentation gap regarding:

Language representation using multiple ISO standards in Alma CZ bibliographic records following the February 2026 enhancement