Google has recently revised its rel canonical documentation to provide clarity on how the platform handles the extraction of rel canonical annotations. This clarification aims not to signify a shift in Google’s approach to handling rel=canonical annotations, but rather to explicitly articulate the process by which Google processes them.
Canonical Link Relation – RFC 5988
Google’s documentation has consistently referenced RFC 5988 as the standard it adheres to regarding the usage of the link relation canonical. This RFC, published by the Internet Engineering Task Force (IETF), outlines specifications for various Internet and networking technologies, particularly those related to the HTML rel link attribute.
In the context of HTML, an element serves as a fundamental component of a webpage and can be enhanced with attributes. In the case of the Link element, it is modified by the Rel attribute.
RFC 6596 further delineates the rel link attribute:
“RFC 5988 specifies a method for defining relationships between links on the web. This document introduces a new type of such a relationship, ‘canonical,’ to designate an Internationalized Resource Identifier (IRI) as preferred over resources with duplicative content.
…Common implementations of the canonical link relation include specifying the preferred version of an IRI among duplicate pages created through the addition of IRI parameters (e.g., session IDs) or designating the single-page version as preferred over the same content fragmented across multiple component pages.”
In essence, the canonical link element indicates when another document is duplicative and identifies the preferred original. These parameters serve as the basis for Google’s processing of the canonical link element.
Changes To Canonical Documentation
The alterations made to the Search Central Documentation were specifically targeted at rel=”canonical” link annotations that extend beyond the typical scenario of specifying duplicative documents, along with some minor adjustments to the page content.
Google made the following modification:
“Google supports rel canonical link annotations as described in RFC 6596.”
The change consists of adding the word “explicit”:
“Google supports explicit rel canonical link annotations as described in RFC 6596.”
Although this change may appear insignificant, it actually serves as the focal point of the documentation update by explicitly stating that Google aligns with the standards outlined in RFC 6596.
Additionally, a completely new paragraph was added.
The new paragraph states:
“rel=”canonical” annotations suggesting alternate versions of a page are disregarded; specifically, rel=”canonical” annotations with hreflang, lang, media, and type attributes are not employed for canonicalization.
Instead, it’s recommended to utilize appropriate link annotations to denote alternate versions of a page; for instance, link rel=”alternate” hreflang for language and country annotations.”
In essence, it advises against using “canonical” to specify elements that are not duplicative webpages, such as pages in different languages or media types, and instead suggests using “alternate” for such purposes.
This clarification doesn’t signify a change in how Google handles or ignores canonical or alternate link elements.
As explained in Google’s changelog documentation:
“Clarifying the extraction of rel=”canonical” annotations
What: Clarified that rel=”canonical” annotations with certain attributes are not used for canonicalization.
Why: rel=”canonical” annotations assist Google in determining the canonical URL among duplicates. Adding certain attributes to the link element alters the annotation’s meaning to indicate a different device or language version. This is solely a documentation update; Google has consistently disregarded these rel=”canonical” annotations for canonicalization.”
Original news from SearchEngineJournal