Table 6 A summary of comments about the nature, perception, and delivery of communications made by volcanologists to end-users relevant to quality standards for contextualisation

Independent, neutral, objective

These are wonderful ideals/aspirations worthy of pursuit. They are difficult to achieve in practice and demonstrate objectively (4, 15, 27).

They can easily be lost by individuals being ‘welded’ by training and/or practice to certain conceptual or theoretical models (15).

Expert elicitation processes can assist objectivity “because the big cheeses cannot just override the other people who might have very good arguments” (15)

Neutral, Value-free, Recipient/Outcome centric

There are occasions when value free neutrality is not possible and appropriate, and it is better to nuance volcano level changes to produce an apparent gradual escalation of unrest with the objective of assisting risk managers and preventing unnecessary public alarm (28).

Societal context should not influence hazard analysis at all (12, 17).

If analytical processes are always constrained by time, resources etc., they should be focussed on selected areas based upon any acquired knowledge of ‘exposure’ and ‘vulnerability’ risk variables (12, 19, 24)

Analytical processes cannot ignore context as scientists must be realistic and recipient-focussed. Scientists must act in a way that helps risk decision makers make informed decisions (3).

A hazard assessment merely starts a conversation with other stakeholders (19).

Commercial, ideological and religious interests cannot be ignored if scientists are to be truly recipient focussed (24). “You can’t ignore them [commercial interests] nor can you…completely detach from the political system in which you’re operating because…there are going to be political realities which constrain what you can do and… there are going to be practical compromises one comes to because otherwise your advice is going to get ignored so you can’t afford to be too prissy” (4).

Balance

It is important to record and communicate not only the view of the majority but also the opinions of dissenters (15, 28).

Understandable

Communications must be recipient focussed, in content and form, and respectful of distinct cultural/community contexts (1, 10, 19)

Good practice

This issue is not straightforward because it raises the complex issue of quality - in particular, values and thresholds for practice quality and practitioner competence.

Views were expressed that quality of practice is: (1) context dependent; (2) dynamic, as it evolves as a function of time; (3) linked to the values of openness and transparency; and (4) related to stakeholder trust.

Peer review

Obvious practical difficulties arise during emergencies, but PR would be more practical if one separated the information being used (the data, analysis, models, etc.) from the advice-giving process itself.

It is important to differentiate between supplementing the expertise/experience of one group of advisers by the informal means of using external competent resources (e.g. a review or second opinion by email or by phone) and true formal PR by independent anonymous reviewers (4).

Provider centric

Recipients are annoyed by words in communications that are included merely “to cover” the scientists providing analysis (24).

At worst this approach is “this is my programme [meaning in this context analysis], this is what it produces, take it or leave it” (1)

Confidence/Trust building

Trust must be built and, once built, you don’t need to worry about it. It can be built by listening to, and working closely with, interested communities and, in particular, commercial and religious stakeholders (24).

It comes from ‘dialogue’ and it is important to have direct communication with religious, commercial and political communities about what ‘impartial’ scientific analysis meant “in their context” (24).

Trust may be linked to ‘independence’. “Independence from…local government is absolutely critical because…[it] is related to trust…If the scientists are increasingly seen as an instrument of government, the trust they have as being independent is…dissipated” (15).

Delivery -Timely, one source, authoritative

“I think that the legal pressure after the L’Aquila trial urged us a lot to write down best practice” (Int. 14).

Sometimes, irrevocable deadlines, which were unnecessary or unreasonable, were imposed and adhered to. Often “a degree of flexibility” would permit something less “half-baked or incomplete” and with less “mistakes” (4)

One source is not always the correct approach. “There are a lot of people saying…there should only be one view, one version. I think that is to underestimate the intelligence of politicians in the public eye. It avoids conflict. It avoids ambiguities but sometimes [more than one source] can serve a purpose”. They opined that, if there are truly two differing schools of thought, expressing them through a sole source may “not necessarily” be the correct way (1).

Written

A ‘written’ communication is important for multiple reasons. The reasons that emerged were (1) to assist openness and transparency; (2) to assist auditing; (3) to prevent misunderstandings and misquotations; and (4) to assist proof of delivery.

Graphics and scientific terminology

The use of graphics provoked a range of views. Whilst some favoured a range of communication forms, including graphics, most the participants accepted that much depended upon the utility attached to graphics by those who receive them.

Terminology “should be as simple and straightforward as possible. You can convey a very complex geological argument using very simple [terms].” (12). Glossaries of the most important volcanic and scientific terms have been used successfully. Continued and frequent use of correct geological terms is important but further relationships will assist inculcation (1).

Clarity, precision and confidence

‘Clarity’ of expressions regarding the likelihood of hazard onset was more important than their ‘precision’. One interviewee (Int. 9) commented, “Once it [the likelihood of hazard onset] is over the threshold, it [precision] is less important”. One interviewee (28) referred to the fine-tuning of precision (the difference between a 65% and 67% chance of a volcanic hazard onset) and added, “Who cares?” Another (12) stated, “Who cares. It does not matter…[fine-tuning] is meaningless. It’s easy to get bogged down in these things.”

Expressions of likelihood are a minefield and a range of solutions was suggested. At one extreme, only numbers should be used without any qualitative narrative. At the other extreme, all expressions (numbers and/or narrative) should be user-focussed and any narrative expressions should be very carefully defined reflecting the different requirements of different audiences. Sometimes locally-calibrated risk analogues are useful. Even the use of words of common usage may be challenging since sometimes common words are used by scientists in a very narrow and technical way.

Confidence (i.e. variability due to limited/lack of knowledge, etc.) was linked by one interviewee with the qualities of ‘openness and transparency’, ‘written’ format and ‘balanced’ content. For others, this difficult concept directly impacted on trust and, accordingly, a recipient focussed approach was needed to reflect cultural and other issues, particularly if qualitative expressions were to be used. The reasons for doing so included: (1) to avoid giving false hope; (2) to identify issues related to resource constraints; (3) to mitigate managerial risks by providing back coverage; (4) to build trust based upon transparency and openness; and (5) to reflect both the quantity and quality of the available evidence.

Tweets

The use of tweets and similar modes of communication provoked a wide range of observations. Some participants warned of unintended over-reactions and consequences, and perceived difficulties of orderly coordination and recipient targeting. Others recognised that a wide range of communication recipient-focussed methods (including highly visual methods, 3D models and videos, and public talks) must be considered and that in some locations and cultures, but by no means all, social media was already very important.

*The authors readily acknowledge that social media practices have greatly advanced since the survey and interviews were conducted.