The reasoned feedback loop

The reasoned feedback loop is central to human progress. Feedback loops are everywhere, but the key element introduced by people is the Reason step. It is core to the way that science works, and engineering, and good (I wish I could think of a better word here), legal, moral and political systems.

In such loops things are in some state, which is examined, and reasoning is applied to do something, to get to a new – intended to be better – state. This seems very abstract, so I will supply some examples to show what I mean. I will also point out where access to information is important in this.


To many non-scientists, the role of a scientist is to know things, but real science starts with not knowing something, but wanting to find out. Scientists start off not knowing, for example if there is a connection between smoking and cancer, or where the energy that powers the sun comes from. They perform experiments, or apply statistical tests, and reasoning, and the end state is an increase in human knowledge.

The success of this process depends on open sharing of the information and reasoning used make the new discovery. Usually these are published in scientific journals, for fellow scientists to see if they can reproduce the results, and examine the reasoning.

This process is international, as spreading good knowledge helps everybody, thus Chinese scientists published an article on ‘A pneumonia outbreak associated with a new coronavirus of probable bat origin‘ in the scientific journal Nature on 3rd February 2020, which is available to anyone to read (note that Nature is normally a subscription publication – that is how they make their money, but they make some articles Open Access). The link to the article came from the excellent Medical Student COVID-19 Curriculum which itself is being improved regularly.

Accurate measurement and good data is essential to all science, and in medicine this is particularly the case, which is why I support the Cochrane foundation – which promoted evidence based medicine, and the All Trials campaign, which pushes for the results of all drug trials, not only the favourable ones, to be made available.

Applying the Reasoning stage is particularly important in the face of a global pandemic. Denying it exists, or humanizing it leads to worse outcomes. An epidemic is the bad kind of a positive feedback loop leading to the bad outcomes, and rationally applied strategies can push towards the good outcomes, as described in ‘Is Coronavirus a Catastrophe

The Law

I shall use the British system here, as it is the one I am most familiar with, and because – despite it’s flaws – it is the result of many people over a long time trying to do The Right Thing.

The Law is not perfect, change can be frustratingly slow, and implementation often fails to match the ideal, but potentially it uses the same process of a reasoned feedback loop as science does.

British Law is made by Acts of Parliament, proposed, discussed and voted on by elected MPs. These discussions are publicly available in Hansard. They are not the most exciting reading, but these parts at least are public.

When there is doubt about the meaning of a law, this is decided by the Court system, refining this through the appeals system until a final judgement is reached. To make these judgements the lawyers use Hansard to try to work out what Parliament intended by the law, and the judgement of previous courts (precedent), to try to make the law as fair as possible. Most of the body of ‘case law‘ is in legal libraries, not published on the Internet, but, for example the British and Irish Legal Information Institute does make many cases available.

Parliament should (and usually does) take existing case law into account when passing Acts which replace previous laws.

Systems where knowledge important to bits of this process are hidden from wider view tend to work less well, as there is more chance that some key information will be missed. If the information is not available to decision makers (for example if their primary source of information are biased lobbyists – and there is no way to review the accuracy of what they have been told) they will make poorer decisions.

The Internet

The Internet, the core part, which should be distinguished from the things which run on top of it, is another example of the effectiveness of open decision making and transparency being used to drive progress. The ‘laws’ of the Internet are a set of documents called Requests for Comments, and they most important of these are published by the Internet Engineering Task Force (IETF). The final versions of these can be found on the web site of the RFC Editor. Those which are Standards are produced by IETF Working Groups, and the discussions which lead to the final documents are openly available (and open to public contribution).

These standards are not developed in isolation, and may have major contributions from companies who are otherwise commercial rivals, but come together to ensure the best possible standards. For example RFC8460 has editors (the people responsible for the final wording of the document), from both Google and Microsoft.

In general for something to be a Standard there must be at least two interoperable implementations.

The guiding principles of the Internet Architecture Board (IAB) an guiding committee of the IETF, is that ‘The Internet is for End Users‘ making the ethical framework explicit.

Target Shooting

When I was a navy cadet I learnt to shoot a rifle, and continued to shoot a little at the Rifle Club when I was a University. On starting work at Harwell, I continued to shoot pistols, a little, at the the range on site. This as also used by members of the UKAEA Police, who were armed, unlike most police in the UK.

In the book Master and Commander, and other books of the Aubrey Maturin series, much of the success of Jack Aubrey is due to his focus on naval gunnery and his insistence that his gun crews practice regularly. He would have read Benjamin Robins‘ book ‘New principles of gunnery’ (which incidentally was translated into German by Leonhard Euler, the Swiss mathematician)

Target shooting is an illustration of the feedback process. A first shot is taken, which will probably not be perfect, and but the next shot should be closer to the target. A reasonably good marksman will not be too far off on their first shot and then improve until at some stage the scatter of their shots is essentially random within some grouping. An excellent shot will start better, and their grouping will be smaller. With a lot of practice it should be possible for anyone to improve.

Similarly, for example, scientists trying to predict the impact of various measures on Covid case numbers, or how they relate to hospitalisations or deaths are unlikely to get their predictions perfect on their first attempts, but should be able to revise their predictions to improve them. This improvement process should be open, and scrutinised to ensure that it is working.

If the gun crews on Jack Aubrey’s ships were only allowed a single practice shot, and if they missed they were to be replaced by a different crew they would never improve. They were also working as teams, and the teams improved together, and also competed with other crews on the same ship, and learnt from each other.


This article reflects on and in some cases restates some of the points in Diversity and Regulation.

In the spirit of the article I welcome thoughtful feedback, either through comments, or via Federated Social Media to

Leave a comment

Your email address will not be published. Required fields are marked *