One of the UK’s leading professional bodies has amended its Code of Conduct after adopting in full a series of recommendations made by the Committee for Academic Freedom, following a constructive process of engagement over concerns that its harassment provisions inadvertently risked restricting lawful expression.
Founded in 1839 and granted a Royal Charter in 1866, the Royal Microscopical Society (RMS) is an international learned society with nearly 2,000 individual members and 63 corporate members working across microscopy, imaging and flow cytometry. It also publishes the Journal of Microscopy, the world’s oldest scientific journal dedicated to microscopy, and in 2025 hosted more than 30 events.
Bringing together such a large professional community inevitably requires rules governing how members interact with one another. The RMS’s Code of Conduct accordingly sets standards of behaviour for members and non-members participating in Society activities, including provisions intended to prevent bullying, discrimination and harassment.
Our concern was not with the existence of conduct rules per se. It is, after all, entirely right that professionals are held to high standards of conduct.
Difficulties arose, however, because the Code contained two formulations which were difficult to reconcile. On the one hand, and commendably, it stated that the Society “cannot police thoughts” and wished to create an environment in which a range of opinions could co-exist respectfully and professionally. Yet the very next sentence stated that, where “personal beliefs” were “enacted in discriminatory comments, behaviours or correspondence”, the Society would act under the policy. The difficulty is that comments and correspondence may themselves be vehicles for expressing perfectly lawful opinions or beliefs, even where someone else experiences that expression as discriminatory.
The same uncertainty appeared elsewhere in the Code. Its harassment provisions encompassed verbal and non-verbal conduct and gave as examples “abusive spoken or written words”, “offensive emails, tweets or comments on social networking sites”, physical gestures and “jokes”. There’s no doubt that any of those things may amount to harassment or some other form of sanctionable misconduct. But the unresolved question was what threshold governed the movement from lawful opinion to sanctionable conduct. In our view, the Code did not provide a sufficiently clear and legally defensible basis for making that distinction.
This was in part because, while the Code reproduced the broad definition of harassment contained in section 26(1) of the Equality Act 2010, it did not reproduce the safeguards in section 26(4) governing how the “effect” of alleged harassment is to be assessed.
To be sure, the Equality Act’s prohibitions on harassment apply only within the specific contexts defined by the Act, and not to every interpersonal disagreement that may arise within a professional society. But the RMS Code had itself adopted Equality Act harassment concepts, while decisions taken under it could affect membership, access to Society events, professional registration or other benefits. In those circumstances, we thought it prudent for the Code to reflect the statutory threshold with care.
Under that legislation, unwanted conduct related to a protected characteristic can amount to harassment where its purpose or effect is to violate someone’s dignity or create an “intimidating, hostile, degrading, humiliating or offensive environment” for them. But where the allegation turns on the effect of the conduct, the complainant’s perception is only one part of a three-limbed test. The Act also requires consideration of the “other circumstances of the case” and whether it is reasonable for the conduct to have had that effect.
Those contextual and objective safeguards are precisely what prevent harassment law from collapsing into a pure test of subjective offence, potentially dragging routine disagreements, clumsy jokes, awkward seminar or conference exchanges, or lawful expressions of protected philosophical belief – including gender-critical belief, anti-Zionism and opposition to critical race theory – into its orbit.
The practical importance of those safeguards was illustrated in Richmond Pharmacology Ltd v Dhaliwal, where Underhill J, as he then was, observed that “dignity is not necessarily violated by things said or done which are trivial or transitory” and that, while it is important to be sensitive to the hurt caused by discriminatory comments or conduct, it is “also important not to encourage a culture of hypersensitivity or the imposition of legal liability in respect of every unfortunate phrase”.
As drafted, however, the RMS framework risked doing just that, inadvertently giving members the impression that offence, distress or disagreement could do too much of the operative work. It did not make sufficiently clear that decision-makers must also consider factors such as the setting of the exchange, the relationship between the parties, whether the conduct was targeted or persistent, whether it formed part of lawful professional discussion, and the seriousness of the language used.
In turn, as we pointed out, this risked generating over-reporting, uncertainty in decision-making, inconsistent enforcement, chilled professional discussion and contested findings of misconduct reached on a lower or more subjective basis than the civil law itself would recognise.
That did not mean members should be discouraged from raising concerns. But there is an important distinction between conduct capable of giving rise to a complaint and conduct ultimately capable of satisfying the threshold for harassment. Making that distinction clear benefits complainants and respondents alike by giving both a shared understanding of the standard against which a complaint will ultimately be assessed.
And there are also benefits for organisations themselves. Recent case law has shown that disciplinary or investigatory action taken in response to the expression of protected philosophical beliefs may itself give rise to claims under the Equality Act. Getting the threshold right is therefore a matter of prudent legal risk management.
For those reasons, CAF proposed a series of targeted amendments designed to put those missing safeguards into the operative provisions of the Code, while preserving the Society’s entirely legitimate anti-harassment aims.
Among other changes, we proposed that decision-makers should, in addition to the perception of those affected, consider the context of the conduct, the nature of any professional or scientific discussion involved, all other relevant circumstances, and whether it was reasonable for the conduct to have had the alleged effect. We also proposed making clear that examples such as offensive communications or jokes would amount to harassment only where, having regard to their context and seriousness, the relevant threshold under the Code – which, under our proposed amendments, incorporated the relevant section 26 safeguards – was met.
The Society’s response was exemplary. Its Chief Executive, Sali Davis, thanked CAF for what she described as our “very helpful knowledge and advice” and referred the proposals to the RMS Executive Board. Following its consideration of the changes, she confirmed in August that the Society had been “able to adopt all of your suggestions”, and that the revised Code would be shared with its scientific sections during the remainder of 2026.
The result shows what can be achieved when concerns about freedom of expression are raised precisely and organisations are willing to engage with them constructively.
And the intervention itself began with a tip-off from a CAF member who had encountered the Code and raised concerns about its implications for lawful expression. Without that initial contact, the issues we identified might never have come to our attention.
This kind of work often takes place out of public view, but it is becoming an increasingly important part of CAF’s wider mission. Protecting academic freedom does not just mean defending scholars after investigations, cancellations or disciplinary processes have begun, but scrutinising the internal policy architecture that shapes how lawful speech is interpreted, reported, stigmatised and escalated before such processes ever begin.
We are therefore keen to hear, in confidence, from staff, students, members of learned societies or others in the academic community who have encountered similar concerns.
In particular, we are interested not only in public-facing harassment or conduct policies, but also in mandatory training modules, EDI courses, staff guidance, complaints procedures and third-party materials hosted behind institutional password systems.
If you are worried that your organisation’s policies or training materials risk encouraging an over-broad approach to harassment, protected belief or lawful professional disagreement, please get in touch at manager@afcomm.org.uk.
Confidentiality will, of course, be respected, and complete anonymity can be maintained where requested.