The UN may NOT be wrong!

The UN may NOT be wrong!

This is not an opinion reversal !

My 'The UN is wrong' blog post from a couple of weeks ago puts the case against the proposed UN database of chemical classifications. It said if standardised hazard classifications exist they will be copied everywhere and go stale everywhere - perhaps forever.

However there is a way for UN standardised hazard classifications to be successful in the long term. 

There is absolutely nothing wrong with standardised classifications. Experts in this realm are scientists and will change their opinions if new evidence is sufficiently persuasive. That means classifications will change.The system used MUST welcome that change.

This is the affirmative case:

  1. Scientific consensus in a standardised classification is gold. Consensus is peer review on steroids. We love it.

  2. Only in an open and fully integrated system can standardised hazard classifications have real value

The affirmative summary

There are only so many experts and there are billions of substances. The system for the proposed UN standard classifications database MUST be open (but not binding) for ALL experts. 

In other words, ALL substance manufacturers must be able to document their substances in the same system so their in-house and/or contracted experts may develop their own classifications via the literature and/or laboratory research. 

The same system MUST be able to classify mixtures according to GHS specifications and formulae.

And the same system MUST also produce SDSs (and labels) for all jurisdictions based on the hazards classified and signed off by an expert.

Such a system also MUST be commercially viable in the long term - independent of the UN.

And if all this exists the UN might not need to make the effort anyway!


Proviso #1 - Non-binding

The UN proposed standard classifications are specified as non-binding. Their excellent reason is that different experts may classify a substance differently - and equally correctly!

Proviso #2 - Experts are responsible

The expert(s) involved in classifying hazards MUST take responsibility for their classifications. They need to rely on the evidence available at the time of the classification. The system should let them link such evidence to each substance.

Proviso #3 - Classifications and SDSs in the same system

SDSs based on any classification MUST be automatically invalidated when that classification changes. This means the classification and the SDSs MUST be in the same system.

Proviso #4 - See Proviso #3

If the classification and the SDSs cannot be in the same system they MUST be in systems which are linked autonomously. There MUST NOT be any human triggering of updates to be forgotten or lost when employees move on without being replaced.

Proviso #5 - Notify updates

When a classification changes, the expert(s) involved need to decide if the change is significant or routine. This means change to workplace practices will or will not be required. If the change is tagged as significant, the system MUST reliably notify all organisations relying on that classification.

Summary so far

Standardised classifications[1] are a Good Idea(TM) but if implemented with last century's technology will definitely be a waste of effort and definitely have long term bad consequences.

Data and classification in the same system

An angle we skipped in the Provisos is that classifications depend on evidence. Therefore substance data must be in the same system. See Proviso #3 and #4 above. Same principle different angle.

Data and evidence connected

The evidence - meaning the research - must be connected to the substance. It is important that data are substantiated. So the substance in the system must have links to the evidence.

It should NOT have copies of the evidence. Copies quietly go stale and no-one notices. It is still there justifying a classification made last century. Oh no. The system the UN should use must have a method of updating evidence for each substance with classified hazards. There are many ways of automating this. Nowadays.

Different experts in the same system

Different scientists must have open access to the standardised classifications[1] so that consensus is possible far into the future. Otherwise it becomes a black box and loses trust. Think about the success of Wikipedia. Different experts must be able to link a substance to other evidence.



[1] The word "Standardised" here means "Scientific consensus" on the classification of hazards of a given substance.

It most definitely does not mean agreed (consensus) hazard classifications for a bunch of substances having the same name but produced by a number of different companies - at the behest of a regulator for the convenience of bureaucrats, lawyers and corporate marketing executives.

Current rating: 5