Appearance
Citation Systems and Persistent References
ISO 690 defines five citation systems — different methods for linking a mention in the text to its corresponding entry in the reference list. All five share the same reference list structure; they differ in how in-text citations are formatted and how the reference list is ordered.
Reference List Ordering
Regardless of which citation system is used, ISO 690 defines rules for alphabetical reference lists:
- A creator's own writings precede resources they cited or edited
- Single-creator entries precede multi-creator entries beginning with the same name
- Multiple entries by a single creator are ordered chronologically (earliest first)
- Multi-creator entries with the same first creator follow single-creator entries
- Corporate creators are alphabetized by the first significant word
- Alternative orderings are allowed if applied consistently
Special cases allow title-first ordering (for classified lists), area-first ordering (for maps), or title-first ordering (for films and audiovisual materials).
1. Name-Date (Harvard) System
The most widely used system in the sciences and social sciences.
In-text: Creator surname + year in parentheses. When the name occurs naturally, only the year is parenthesized. Page numbers may follow the year.
(Crane 1972)
Crane (1972) argues that...
(Crane 1972, pp. 23–25)Same creator + same year: distinguished by lowercase letters — (Crane 1972a), (Crane 1972b).
Reference list: Alphabetical by surname, year immediately following.
Example reference list entry:
CRANE, D., 1972. Invisible Colleges. Chicago: University of Chicago Press.2. Numeric System
Common in engineering, physics, and standards documents.
In-text: A number (superscript, parentheses, or brackets) referring to the entry in order of first citation. Once assigned, a reference keeps its number throughout.
sciences [26]
Stieg [13, p. 556]Reference list: Arranged in numerical order of first citation.
3. Named Tag System
Author-chosen alphanumeric labels serving as both citation markers and reference list keys.
In-text:
[INV]
[ISO44001]
[HISTNEEDS, p. 556]All citations of the same resource use the identical tag throughout.
Reference list: Alphabetical by tag name.
This system is particularly natural for standards documents, where references are cited by their document identifier (e.g., [ISO 19115-1:2014]).
4. Running Notes (Footnote/Endnote) System
Used in the humanities and legal documents.
In-text: Superscript number referring to a footnote or endnote, numbered in order of appearance.
Notes may contain full references or short-form references ("Crane, op. cit., p. 23"). Multiple sources may be cited in a single note. Earlier-cited resources may be referenced by their note number.
The first note should explain all abbreviations used (ibid., op. cit., loc. cit., etc.).
Running notes typically do not use a separate reference list — all bibliographic details appear in the notes themselves.
5. Implied System
The document identifier itself serves as the citation, without additional markers.
In-text: The identifier appears directly in running text:
As specified in ISO 44001:2017...
A successful example can be found in IETF RFC 3113.Reference list: Logical order by document identifier, optionally grouped by category. This system is common in standards, where publications are cited by their PubID.
Exclusive implied list example:
ISO 8601-1:2019, Data elements and interchange formats...
ISO 8601-2:2019, Data elements and interchange formats...
CC 18012, CalConnect...
IETF RFC 3339, Date and Time on the Internet...
NIST SP 800-90B, Recommendation for the Entropy Sources...The implied system can be combined with other systems: implicit references may follow the other system's ordering, be grouped separately, or appear in a dedicated bibliography.
Relaton's Support for All Five Systems
relaton-render supports all five citation systems through configurable templates:
| Flavor | Primary System | Notes |
|---|---|---|
| ISO, IEC | Named tag / Implied | References cited by document identifier |
| IETF | Named tag / Implied | RFCs cited by number: [RFC 8446] |
| ITU | Named tag | Recommendations cited by identifier |
| NIST | Name-date / Numeric | Depends on NIST document series |
| BIPM | Numeric | Brochure references numbered sequentially |
The underlying Relaton data model is the same regardless of rendering — a single BibliographicItem can be rendered in any citation style.
Persistent References to Web Resources
ISO 690 Annex B addresses a critical problem: link rot (the link returns 404) and content drift (the content has changed). URLs are not reliable long-term identifiers. The standard provides four strategies:
1. Persistent Identifier (PID) Systems
PIDs (DOI, Handle, ARK) provide registration and resolution services. They protect against link rot but may not protect against content drift — a DOI for a newspaper homepage resolves to the current (changed) content.
PIDs should always be used when available. However, if a specific version is needed, a web archive URI should be used instead.
2. Permanent Link Systems
Services like PURL and Perma.cc issue persistent URLs that redirect to the current location.
URL shorteners (bit.ly, etc.) shall NOT be used because they lack persistence, do not function in web archives, and have unknown support lifespans.
Permanent links may be used only if: the resource has one, the system is sustainable, and the reference also includes an archived copy and viewing date.
3. Web Citation Services
Archive a resource and provide an identifier. Example:
Available from: https://www.theguardian.com/...
Archived copy available from: WebCite [distributor],
http://www.webcitation.org/5Kt3PxfFl
[archived 2006-12-04T19:19:45Z] [viewed 2019-02-28].References using web citation services should also contain the original URI for searching other archives if the service fails.
4. Web Archives
Services like the Internet Archive harvest and preserve web resources. Key guidance:
- An archived copy should be cited if: the resource is no longer available, the resource is dynamic, or there is a need to ensure no link rot or content drift
- Choose the copy that best fits the purpose (most liberal access, most sustainable)
- If no archived copy exists, create one — Internet Archive for free access, national web archives for legal deposit
- Internet Archive applies
robots.txtbans retrospectively — a previously archived page may disappear - Only the archival copy actually used should be cited, not copies from other archives
- If accuracy matters, specify relevant component parts separately (images may not be harvested simultaneously with text)
Relaton's approach: The model supports all four strategies through typed URIs (uri.citation, uri.xml) and date tracking (date.accessed). The auto-fetch mechanism resolves PIDs to retrieve structured metadata.