[SystemSafety] Fetzer

Peter Bernard Ladkin ladkin at causalis.com
Thu Jun 20 17:56:40 CEST 2019


On 2019-06-20 16:17 , Tom Ferrell wrote:
> Perhaps you meant this as a rhetorical question...

I read it as Martyn asking us to try to explain why certain kinds of events/results are accepted as
evidence for dependability and other kinds not, even when they contribute substantially to
dependability properties.

> People reject what they don't understand.

There is an argument to be made that they don't understand "... claims based on testing and process
conformance ..." either, else they wouldn't support them when they are unsupportable.

I'd like to modify your suggestion. I think it is an epistemic issue: people reject what they know
or believe that they don't understand. And many believe, truly or falsely (and often falsely), that
they understand testing and they understand how process models contribute to dependability.

But I also don't think this is the main factor in play. In standards committees, people are most
obviously representing what they take to be the interests of their companies, while paying at least
lip service to "scientific standards". (I say "what they take to be .... " because arguably the best
interests of their companies are served by supplying kit which objectively conforms to the
dependability criteria they claim it conforms to, rather than making higher claims than those which
they can scientifically substantiate. And I say "at least lip service" because I have encountered
people who are manifestly trying to bulldoze things which don't fit scientific consensus through
committees, and on the other hand there are people who will only support good science, but cannot
always tell - and know they cannot always tell - what is good and what is questionable about
specific methods.)

This interest-representation is legitimate in that it is explicit and understood that companies and
their personnel will contribute when they see an advantage to doing so. Committees support claims
based on testing and process conformance because that is how their companies persuade customers to
buy their kit. Some of that kit is controlled by assessors, and of course assessment companies sit
on the standards committees also. But participants are all aware that the rising tide lifts all boats.

The obvious weakness in this is poor (engineering-scientific) quality control. That is why John
Knight and I have both argued for independent assessment of draft standards. I argued along the
lines of the EU-project model, because I have experienced how that model has improved project output.

I have been involved in a number of (persistent) attempts to improve poor draft standards. I have
experienced mostly that such attempts fail.

PBL

Prof. Peter Bernard Ladkin, Bielefeld, Germany
MoreInCommon
Je suis Charlie
Tel+msg +49 (0)521 880 7319  www.rvs-bi.de





-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: OpenPGP digital signature
URL: <https://lists.techfak.uni-bielefeld.de/mailman/private/systemsafety/attachments/20190620/2db66b4a/attachment.sig>


More information about the systemsafety mailing list