Now let us turn our attention to normative references and the most frequently asked questions and commonly experienced problems.
First,
let's start with the basics:
What is a
normative reference?
A normative
reference is a document to which reference is made in such a way as to make it
indispensable for the application of the standard. This therefore assumes that obtaining
any document which is normatively referenced is essential for the proper
implementation of the standard.
Normative
references have to be referred to normatively in the body of the document at
least once. See Table H.1 in the Internal Regulations Part 3 for precise guidance on how to express requirements.
To
differentiate, an informative reference is either a document which is directly
referenced in the standard, but the use of which is not indispensable for the
application of the standard, or a document that the drafting body used as an
informal source for the drafting of the text. Informative references should be listed
in the bibliography.
The key question to ask when tackling normative references is whether
the document you are referencing is in fact indispensable to the application
of the standard and is it directly referenced at least once in the body of
the text? Is it normative or informative?
|
How
and where should I list normative references?
Here
is an example of how a normative reference should appear in Clause 2:
EN 50539-11, Low-voltage surge protective devices – Surge
protective devices for specific application including d.c. – Part 11:
Requirements and tests for SPDs in photovoltaic applications
|
When does a normative reference have to be dated?
References
need only to be dated when they are being used to refer to a specific element (e.g.
clause, subclause, annex, numbered list entry), formula, table or figure in the
referenced document. This is because a new edition of a standard may not have
the exact same order of content as its previous version.
Where both normative and informative references are undated then the
latest version of the referenced document, including any amendments applies.
It is
essential that the normative references are as up-to-date as possible and that
the latest version is used so that the published text remains state of the art.
It is also essential that any document which is normatively referenced is publicly available and not at draft stage.
In
certain special cases, references may have to be dated, even though no specific
element from it is mentioned. For example, in some cases deliverables are required
to meet the Essential Requirements of New Approach Directives, in which case normative
references which appear in a clause which is listed in the Annex ZA may need to
be dated in Clause 2.
This
occasionally calls for a case-by-case approach during editing and a close
collaboration between the editor, the drafting body and other stakeholders –
e.g. Consultants.
If you
have any queries about the document you are working on then feel free to drop
us a line at production@cencenelec.eu.
Where can I find
the information regarding the most recent version of an EN and its title?
You
can find all the necessary information on Projex-Online.
Other helpful sources of information are the Search areas on the CEN website (http://standards.cen.eu/dyn/www/f?p=CENWEB:105::RESET::::)
and on the CENELEC website (http://www.cenelec.eu/dyn/www/f?p=104:107:2583858426572501::::FSP_LANG_ID:25).
What if I have a normative
reference which is not a CEN, CENELEC, ETSI, ISO or IEC standard?
Are
you drafting a CENELEC EN? There is a
CENELEC Guide 10 that has been adopted by the Technical Board and gives the
relevant information: http://www.cenelec.eu/membersandexperts/referencematerial/cenelecguides.html.
Are
you drafting a CEN EN? The responsible Technical Committee has to take a formal
TC Decision. The formatted decision can
be found in Section 2.2 on the following page of the CEN BOSS website: http://boss.cen.eu/reference%20material/FormattedDecisions/Pages/default.aspx
and must be returned to CCMC at: tcminares@cencenelec.eu.
Summary
o
Optional element
o List documents
which are indispensable for the application of the standard
o Date references
if they refer to a specific element in another document, e.g. table, figure…
o Refer preferably
to other European Standards
o Standards from other
organizations can be referenced under certain conditions
o Normative
references must be referred to normatively in the body of the text at least
once in order to appear in Clause 2
o Avoid references
to:
o drafts
(unpublished work)
o documents that
are not publicly available
o documents that
are referred to informatively
o documents that
have served as background material in the preparation of the document
|
Normative
references can be a difficult aspect of drafting standards and the Editors at
CCMC are always here to provide you with support on any queries you might have
throughout the drafting process.
Any questions
about the topics covered in this post? Post us a comment!
And
don’t forget to follow us – next week we will be blogging about terms and
definitions…
Frédéric
Dufour and Sarah Gibson (fdufour@cencenelec.eu,
sgibson@cencenelec.eu, production@cencenelec.eu)
© All CCMC posted information is copyrighted by CEN and/or CENELEC.
Extracts may be used providing that the source is mentioned.
Please note that the CCMC posted information is drawn from a variety of
sources already publicly available on the CEN and CENELEC websites.
Having discussed this subject at great length in IEC and in industry, I
ReplyDeleteagree that the advantages and disadvantages of voluntarily dating
references (i.e. where the rules don't compel dating) are fairly evenly
balanced.
BUT the big advantage for the user of the standard is that they
definitely know which edition of the referenced standard applies, and,
especially, those in industry who do not participate in standards work
are very grateful for this.
In addition, at a higher 'philosophical' level, no committee can be
quite certain that all future editions of an undated standard will be
correctly applicable.
In IEC TC100, as General Maintenance Manager I am trying to get
committees to regularly (annually?) check the references (both dated and
undated) in their standards for continuing validity. It is nominally a
task for the Secretariat, but there seems no reason why the task should
not be divided among a number of *reliable* committee members.
The editors responded and I answered:
>Please be aware though that the use of undated references is all the
>more useful when a text has to be reviewed or consolidated; having all
>the references dated would make the reviewing or consolidating of a
>text far more complex.
I disagree that it is useful; instead, it is risky. By just accepting
the undated references, **without checking that the latest version of
each is still correct and relevant**, the reviewer is in danger of
allowing a standard with a serious error to be published. It might not
happen very often, but it ought not to happen at all, and fate decrees
that it will happen in a safety standard, where it matters very much.
The checking of numerous dated referenced standards to see whether they
have been superseded and if the new versions are still relevant is also
time-consuming, ** but if it is not done, the standard that makes the
references is not in error**.
If numerous undated referenced standards are not checked, the standard
that makes the references is quite likely to be in error.
The editors replied and I responsed again:
>Furthermore, still in the case of a review, all the references have to
>be thoroughly checked, even if they are undated. This approach may be
>time-consuming, I agree, but it has to be seen as a necessary
>discipline to ensure the new version of the text will be "state of the
>art".
This seems to be inconsistent with what you wrote before:
"Please be aware though that the use of undated references is all the
more useful when a text has to be reviewed or consolidated; having all
the references dated would make the reviewing or consolidating of a text
far more complex."
It seems no less complex to review the undated references as the dated
ones. Your statement above seemed to imply that undated references need
not be checked, making the review task easier.