PreprintPDF Available

First Choirs Standing? What risks were taken by choirs returning early to singing during the 2020 Covid pandemic, how were the risks managed and what were the outcomes?

Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

This paper is a work-in-progress preprint. It has not yet been peer reviewed and an updated version of it is now being prepared. It has been made available because of the urgency of the topic and the rapidity with which the field is changing. Abstract The 2020 pandemic caused by the SARS-CoV-2 virus ("Covid-19") resulted in a worldwide cessation of choral singing. Resumption of choral singing took place at varying times and to different degrees according to conditions pertaining locally around the world. In all cases it was recognized that the activity must be risk-assessed and that until the SARS-CoV-2 virus has been eliminated by means such as vaccination, the precautionary principle must apply to any future choral rehearsal or performance.In the absence of a more specific or quantifiable assessment, it is difficult to determine whether a rehearsal or performance should go ahead. However, there are also drawbacks to quantification as probability. Low probabilities do not eliminate risk and may even generate secondary uncertainty. Choirs that resumed rehearsal early in the pandemic devised control measures that, in the judgement of their managers, would result in an acceptably “low” risk. Given the level of scientific uncertainty prevailing at the time, a high level of precaution was applied to the interpretation of “low” or “acceptable” risk. The judgements of the “first choirs to stand” appear to have been vindicated in that no infections associated with their performances were reported. Subsequent scientific research has also vindicated the judgement of the first choirs to stand that airborne infection was a significant risk that should be controlled for. At the time these choirs planned their performances, the World Health Organisation was advising that “no airborne transmission had been reported”.
Content may be subject to copyright.
First Choirs Standing? What risks were taken by choirs returning
early to singing during the 2020 Covid pandemic, how were the
risks managed and what were the outcomes?
Martin Ashley PhD
Editor-in-chief, ABCD Choral Directions Research
Status of this paper
This paper is a work-in-progress preprint. It has not yet been peer reviewed and an updated version
of it is now being prepared. It has been made available because of the urgency of the topic and the
rapidity with which the field is changing.
The 2020 pandemic caused by the SARS-CoV-2 virus (“Covid-19”) resulted in a world-wide cessation of
choral singing. Resumption of choral singing took place at varying times and to different degrees
according to conditions pertaining locally around the world. In all cases it was recognized that the
activity must be risk-assessed and that until the SARS-CoV-2 virus has been eliminated by means such
as vaccination, the precautionary principle must apply to any future choral rehearsal or performance.
Risk is often expressed in general terms such as “high”, “low” or “acceptable”. In the absence of a
more specific or quantifiable assessment, it is difficult to determine whether a rehearsal or
performance should go ahead. However, there are also drawbacks to quantification as probability.
Low probabilities do not eliminate risk and may even generate secondary uncertainty. Choirs that
resumed rehearsal early in the pandemic devised control measures that, in the judgement of their
managers, would result in an acceptably “low” risk. Given the level of scientific uncertainty prevailing
at the time, a high level of precaution was applied to the interpretation of “low” or “acceptable” risk.
The judgements of the “first choirs to stand” appear to have been vindicated in that no infections
associated with their performances were reported. Subsequent scientific research has also vindicated
the judgement of the first choirs to stand that airborne infection was a significant risk that should be
controlled for. At the time these choirs planned their performances, the World Health Organisation
was advising that no airborne transmission had been reported”.
Last Choir Standing was a formulaic TV talent show broadcast by the BBC during 2008. First Choir
Standing has a rather different and altogether much weightier meaning. It refers to the first choirs to
bring singers together to perform in an enclosed space after the global cessation of such activity
resulting from the SAR-CoV-2 pandemic. Though popularly attributed to a wet market in Wuhan,
China, phylogenic analysis has shown ancestry for the SARS-CoV-2 virus throughout 2019, with
evolution culminating in November (Li et al 2020; Andersen et al, 2020). The world did not take the
problem seriously, however, until it became clear that the outbreak in China was spreading globally
during February 2020. In Europe, Italy was the first country to be stricken seriously and went into
lockdown on 3rd March. The now infamous outbreak during a choral performance in Amsterdam
occurred a week later, on the 8th March, with the similarly devastating Skagit Valley rehearsal in
Ashley First Choirs Standing Work-in-progress preprint September 2020
Washington state on 10th March. Many countries imposed national lockdowns soon after these events,
both Germany and the United Kingdom doing so on the 23rd March. From March onwards, choral
singing was largely off the global agenda.
The justification for such severe measures is that the SARS-CoV-2 is a novel coronavirus for which
there is no evolved immunity, and no drug treatment or vaccination programme presently available.
Moreover, the SARS-CoV-2 virus appears to result in an unusual degree of variance in potency and
effect, being apparently unpredictable in how it affects individuals. Effects range between
asymptomatic infection through mild flu-like symptoms to respiratory or multi-organ failure and
death. Longer-term effects specific to singing, including debilitating post-viral fatigue (Williams, 2020;
Moldofsky, and Patcai, 2011) or lasting damage to vocal apparatus, have been identified as requiring
further investigation (Helding et al, 2020). The known risk from asymptomatic carriers or those with
mild symptoms also appears to be higher than first assessed (Havers et al, 2020) and likely to increase
as schools return (Stein-Zamir et al, 2020), though this must be balanced against an even greater risk
of keeping schools closed (DELVE, 2020). The overall level of scientific uncertainty and hence perceived
risk is unusually high (King et al, 2020).
Governments have had little option under such circumstances to exercise the precautionary principle
with significant rigour. Inevitably, interpretations and practices relating to precaution have varied
around the world with the result that, whilst choral singing was prohibited in the UK for four months,
earlier resumption was allowed in other countries. However, until the virus is eliminated, which may
be possible only if an effective vaccine can be found, choirs around the world are going to have to
adapt to a risk-assessment driven process. There is no consensus yet as to how the risk-assessment
process should best be informed and managed. Should there be quantification of probability, or
merely a broad qualitative expression? Should governments or health authorities be the main
assessors, or should there be a degree of devolution to choirs allowed to make their own autonomous
judgements? If so, what knowledge bases should be accessed by choir managers attempting to devise
control measure that demonstrate risk mitigation? This paper takes an empirical approach to the
issue, observing how choirs that have resumed singing have done so and looking for evidence of the
consequences. There is no guarantee that the first choirs to have stood will not face retrenchment if
conditions initially favourable deteriorate. The intended approach is iterative. The first choirs to have
stood will be, as much as is possible, revisited at periodic intervals to document consequences and
changes. It is argued that there is a need for such empirical work as well as the experimental and
randomised control trial approaches that have dominated the science.
National approaches to managing pandemics
Until the situation described above changes, only non-pharmaceutical interventions (NPIs) can be
deployed. The UK government was significantly influenced in the early stages by Ferguson et al (2020)
whose epidemiological modelling suggested two possible approaches, mitigation or suppression.
Mitigation aims to slow but not stop epidemic spread, thereby managing peak healthcare demand and
protecting the most vulnerable. Suppression attempts to maintain a low level of case numbers
indefinitely. Not considered by Ferguson et al is a third strategy, that of elimination. Elimination
attempts to achieve total eradication of the virus, a successful strategy in the case of Smallpox but not
yet in the long running case of measles/rubella (Public Health England, 2019). The suppression strategy
has been followed in the UK on the advice of the government’s SAGE group and is the familiar one
based upon an initial lockdown followed by easing of restrictions. It has been the pragmatic choice of
Ashley First Choirs Standing Work-in-progress preprint September 2020
the majority of advanced democratic economies that on the one hand recognise that mitigation would
overwhelm their health care systems whilst on the other, elimination is impossible by NPI alone.
In the present paper, Norway, Germany and Australia have followed different versions of the
suppression strategy. They do so, however, at a cost. Ferguson et al are quite clear that until an
efficacious vaccine has been proved to work and administered to most of the population, suppression
will have to be maintained indefinitely. Their modelling showed that in order to keep R below 1, a
combination of case isolation, social distancing of the entire population and either household
quarantine or school and university closure are required in perpetuity. Such measures must also be
supported be a reliable “track and trace” system which the UK has been slow to develop. Although
only a stochastic model, this work seems to be proving chillingly prophetic at the present time of
writing when the possibility of closing UK pubs, restaurants and other businesses as a consequence of
schools and universities reopening is being considered.
Stage et al (2020) in a comparative study of schools re-opening across Europe make a similar point to
the one made by the present author about choirs. Where the national R number is low, schools might
reopen, as might choirs. Where it is higher, both schools and choirs create a risk of further outbreaks.
Norway locked down relatively early and completely on the 12th March, closing schools from that date.
A low “R” was attained earlier than many countries, but the Norwegian Institute of Public Health
published the following warning on the 5th May:
The epidemic has so far resulted in a low burden of disease in Norway, but the burden of action is
significant, including both well-documented socio-economic consequences and probable public health
consequences. . . . The overall goal should be that the burden of the disease of the epidemic should
remain low, the health service should not be overloaded, and adverse ripple effects and costs of
infection control measures should be low. To achieve this, the strategy must be dynamic. . . .adjusted
according to the development of the epidemic and knowledge. The core is still hygiene measures, early
detection and isolation of infected and tracking follow-up (and quarantine) of close contacts to the
infected. (Folkehelseinstituttet, 5. mai 2020)
The impact of the Norwegian strategy will shortly be considered through a case study of the Nidaros
(Trondheim) Cathedral boys’ choir.
Some countries did not introduce widespread suppression lockdowns. Some were unable to do so on
account of large, dispersed populations and poor levels of education and communication. This obvious
difficulty may have discouraged the World Health Organisation from advocating measures that would
be unaffordable for poor countries (Tang, 2020). In others, erratic leadership from mostly right-wing
populist regimes seems to have been a common factor. However, Sweden followed a deliberate
calculated policy that perhaps owes more to mitigation than suppression. Considerable trust was
placed in the country’s national epidemiologist, Anders Tegnell.
Tegnell’s more liberal policies were widely misrepresented in foreign newspapers. It was not true, as
some claimed, that there were no lockdown measures. Universities and high schools were ordered to
close on 17th March, whilst cafes, restaurants and bars were confined to table service only from 24th
March. Large outdoor gatherings (50+) were also banned and exhortations to wash hands regularly,
maintain social distance and work from home were issued much as in other countries. Nevertheless,
primary schools did not close and Tegnell was of the view that Sweden’s traditional stress upon free
will and individual responsibility should be respected. He specifically denied that so-called “herd
immunity” was being attempted, citing the need to balance health measures against measures of
economy and mental well-being. The results have made for interesting reading. Compared with its
close neighbour, Norway, Sweden appears to have fared badly. By 6th June, Sweden had recorded
Ashley First Choirs Standing Work-in-progress preprint September 2020
4468 deaths out of a population of 10.18m against Norway’s 237 deaths out of 5.31m. The deaths per
100 000 were respectively 43.2 (Sweden) and 4.4 Norway.
However, by the same measure, the UK had fared worse than Sweden with 58.5 deaths per 100 000
by 6th June. By the 28th July, Sweden was able to report a rapidly falling infection rate with 398 new
cases by the beginning of August as against 767 the previous week and 2530 at the beginning of July.
Reasons for this apparent success will doubtless be put forwards, but as Tegnell reports, the
predictions of the Ferguson et al model for Sweden were not realised. What is of immediate
significance is the “Achille’s heel” of Tegnell’s strategy. As an analysis by Kamerlin and Kasson (2020)
shows, there was a higher proportion of deaths amongst the older section of the population than in
countries with more stringent lockdown, a fact not denied by Tegnell. As will be seen later, this is of
some significance for the choir studied the Stockholm Boys’ Choir.
Finally, the perhaps improbable elimination strategy has been rigorously pursued by New Zealand.
Importantly, the New Zealand Ministry of Health clarify that they are not attempting total eradication,
as in the case of smallpox.
Elimination of COVID-19 (or any disease) means reducing new cases in a defined geographical area, in
this case Aotearoa/New Zealand, to zero (or a very low defined target rate). Elimination is distinct from
eradication. Eradication refers to the complete and permanent worldwide reduction to zero new cases
of the disease through deliberate efforts (eg, smallpox). Eradication of COVID-19 is not possible at this
stage (and may not ever be possible) (Ministry of Health, Government of New Zealand, 2020)
In order to achieve this, the NZ government published early on a four-level alert system:
Level 4: Lockdown, likely the diseases is not contained. Introduced 25th March
Level 3: Restrict, high risk the disease is not contained. Introduced 27th April
Level 2: Reduce, the disease is contained but the risk of community transmission remains.
Introduced 13th May.
Level 1: Prepare, the disease is contained in New Zealand. Introduced 8th June
The strategy appears to have been successful in that by the 8th of June, Covid had indeed been
eliminated, if not eradicated from New Zealand. The country is on standby to revert to higher levels
should the situation deteriorate.
Such has been New Zealand’s success that the UK’s Independent
Sage group has deviated from the official Sage group to advocate an elimination strategy (King et al,
2020). A key argument is that the UK, like New Zealand, is an island and could therefore exclude cross-
border contamination. This seems an unrealistic proposition given Britain’s position as a major
European air hub and is regarded by Tegnell as unattainable. Nevertheless, the Scottish government
has deviated from Westminster by pursuing an elimination strategy, and at the time of writing,
Scotland does indeed appear to be faring rather better than England (566 deaths out of 5.5m
population in Scotland, 46 566 deaths out of 59m, England). The case of the New Zealand elimination
strategy is illustrated shortly by Auckland Cathedral’s music department.
Approaches to risk management
Three fundamental approaches to risk management have been identified as being in use.
Control measures only
Ordinal scales
Since this was written, an outbreak occurred in Auckland that is currently being investigated. It is possible it
entered the country through contaminated freight.
Ashley First Choirs Standing Work-in-progress preprint September 2020
In this section, each is described briefly and assessed for suitability in the practical situation of choir
management by lay people. As will be seen later, the risk management strategies used by the first
choirs standing were largely of the control measures only variety. The discussion will tackle the
question of whether either of the other two strategies would have been preferable as well as
attempting an evaluation of how effective the control measures were.
Control measures only
An example of this approach has been produced by the Church of England and is illustrated in Fig I.
The assumption is that if the controls are all carried out and shown to have been done so in
documentation, then the risk will have been reduced to the lowest practicable level and the person
responsible will have acted reasonably and in good faith.
Figure I
Control measures recommended for church cleaning
A wide range of scenarios is covered, and the controls seem comprehensive. No scientific references
are given though it can be presumed that the House of Bishops has been careful to take appropriate
advice. For example, the advice that there is no need for extra cleaning if the church has been closed
for 72 hours or more is presumably derived from work on surface stability (van Doremalen et al, 2020)
though there is no means of knowing the extent to which work on complicating or confounding factors
such as stability in different conditions (Chin et al 2020) or variations in the viral load of individuals
(Heneghan et al, 2020) has been considered.
Ordinal scales
The person(s) responsible for implementing given control measures does not actually carry out a risk
assessment, they merely carry out a set of instructions and must take it on trust that they will have
Ashley First Choirs Standing Work-in-progress preprint September 2020
reduced the risk acceptably by following the control measures they have been given. When ordinal
scales are used, the risk assessor is involved in making a judgement about how effective the control
measures might be and has the option of demonstrating increased risk mitigation through the
introduction of further control measures. A scale of 1 5 is used to determine the likelihood of a
hazard occurring and the severity of the consequences. The method is a well-established one that has
been in use for some time in the heritage railway industry where safety-critical risk assessments are
required by the Railways and Other Guided Transport Systems (Safety) Regulations 2006 (ROGS). The
present author has long-standing experience of this system in his capacity as safety manager for one
such enterprise. It has also been used in the UK by Music Mark, a corporate member association for
the music education sector.
In this approach a hazard is defined as “anything that has the potential to cause actual harm” whilst
risk is given by a simple computation of the product of how likely the hazard is to occur (L) and the
seriousness of the consequences of its occurring (C). R=LxC. The critical difference between this system
and following given controls is that the computation must be done twice, once before the
identification of control measures (R1) and again after the implementation of those measures (R2). If
there is a clear difference between R1 and R2, the control measures can be deemed effective. The
values of R are also read off against a table of risk acceptability with actions to be taken which range
from ‘no action required but continue to review’ (1-2) to ‘cease the activity immediately’ (20-25). This
type of approach was advocated for choral singing by Spahn and Richter (2020) who stated that
“effective risk management usually requires a precise risk analysis with an associated likelihood of
occurrence and knowledge of the effectiveness of certain risk-reducing measures.”
Spahn and Richter give little practical guidance as to how this is to be done. Figure II below is an
adaptation for choirs by the present author of a system in use on several large heritage railways.
Figure II
Categoric Risk Mitigation adapted from ROGS for Choirs
Figures IIIa and b give examples of the system in use. In Figure IIIa, the assumption is made that the
choir has a mixed age demographic that includes members over 70. It is well established and little
Ashley First Choirs Standing Work-in-progress preprint September 2020
disputed in the literature that there is a strong relationship between age and susceptibility. Ferguson
et al (2020) give an infection fatality ratio of 0.006% for the 10 19 age group. This rises alarmingly to
5.1% for over 70s and 9.3% for over 80s. Only 0.3% of infected 10 19-year-olds require hospitalisation
as opposed to 24.3% for over 70s and 27.3% for over 80s. Of the 24.3% admitted to hospital 43% of
over 70s and 71% of over 80s require critical care, i.e. ventilator treatment (Ferguson et al, 2020: 5).
There can be very little doubt than that if such persons are in a choir, a high number for the likelihood
of infections must be selected. Similarly, a high number for the consequences. In the example below,
4 is selected for likelihood (meaning “likely”) and 4 for consequence (meaning “several hospitalisations
requiring oxygen”). 4 x 4 = 16 which reads as “take urgent action and stop activity if necessary”. The
risk is clearly unacceptable. However, a control measure of “over 70s and persons with underlying
health issues asked not to attend” can be devised and implemented. When this is done, the likelihood
decreases to 2 (“unlikely”) and the consequence to 3 (“all families have to quarantine for 14 days”). 2
x 3 = 6. The risk is just acceptable, but the situation requires constant and regular monitoring with
periodic reassessment.
The choir could, in theory, then go ahead and rehearse together. As will be seen later, this happened
fairly quickly in the countries where the “first choirs stood”, but the UK government, acing on the
advice of Public Health England, took a more risk averse approach, prohibiting all choral singing for
much of the summer.
Figure IIIa
Part of a categoric mitigation risk assessment for hypothetical mixed-age choir
Figure IIIb illustrates another well-known hazard where the evidence suggests that the L number
should be higher than 2. It is almost certain that respiratory mucus particles will fall onto music copies
where they will evaporate leading to some build-up of viral concentration. This method of fomite
transmission has been assumed to be significant, though some authors point out that it is close person
to person contact that is the more likely means of transmission (Goldman, 2020; Somsen et al, 2020).
Should music copies become contaminated, the virus remains viable on a printed paper surface for up
to three hours (Chin et al, 2020). Fomite transmission through a singer handling a copy used by
another singer less than three hours previously might merit a 2 or 3. The method does not eliminate
subjective judgement, but an L number of 5 might well be justified for two singers sharing one copy
because this would bring them into close proximity.
A control of ‘issue personal copies in plastic folders which are kept for duration and must not be shared
or left lying about’ would address both issues and in the example below brings the L number down to
1, meaning that where there is no government prohibition, this consideration alone need not stop the
choir singing.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Figure IIIb
Categoric mitigation for paper music copies
It is important to stress that the numbers 1 - 5 are of limited statistical power because the are ordinal
and not suitable for parametric treatment. Neither are they measures of probability. Whilst it is true,
valid and useful to state that a 2 is quite a lot better than a 4, it would not be true to state that a death
is five times as bad as a singer absent for a couple of rehearsals. There is a tension here with the
requirement set out by Spahn and Richter for a “precise measurement”. They state:
Effective risk management usually requires a precise risk analysis with an associated likelihood of
occurrence and knowledge of the effectiveness of certain risk-reducing measures. At the moment,
however, we do not yet know much about the transmission of SARS-CoV-2, so that risk management
currently means an equation with many unknowns. (Span and Richter, 2020: 32).
An “equation with many unknowns” may be a risk in itself. Ordinal risk assessment scales work on the
principle that the assessors make an “educated decision”. This is unproblematic when the risk is a
well-known one such as falling off a ladder, and the assessor is knowledgeable and experienced in the
industrial use of ladders. The SARS-CoV-2 virus, however, is a novel one and the number of people
who can truly make “educated decisions” embracing all the associated hazards must be very small.
Other objections, such as a failure to allow for frequency of exposure (a choir meeting daily should
not score the same as a choir meeting only once a month) can be overcome by a more sophisticated
calculus but the “educated decision” difficulty remains hence the case for the epidemiological laity
taking control measures on trust as in the Church of England template.
Calculating probability
Risk as a probability can be computed empirically if sufficient statistical data are available. Obvious
examples would include calculating relative risks in transport which can be on a basis such as miles
travelled per casualty. Ample data over many years are available. No such data are available for SARS-
CoV-2. The nearest approximation might be for a previous novel coronavirus, though how useful this
might be is open to question. Equations from which probability might be derived exist or are being
developed for many discreet events and circumstances relevant to different types of hazard
associated with the transmission of the SARS-CoV-2 virus. One such is the Wells-Riley which is well
established in the field of airborne infection.
The Wells-Riley equation is based upon the concept of a quantum of infection, first proposed by Wells
in 1955. Wells’ work was seminal in persuading a sceptical world that viral infections were transmitted
by airborne routes. The “quantum” he proposed is the minimum dose necessary to cause infection in
a host. The task is to model the accumulation, transmission and distribution of quanta in confined
spaces. In 1955, the focus of interest was pulmonary tuberculosis, and this has not gone away since
that disease has not been eradicated (Nardell, 2016). The WellsRiley equation was itself developed
by Riley et al in the context of a measles epidemic (Riley et al., 1978). Measles is known to be highly
contagious through the airborne route. Noakes and Sleigh (2008) pointed out that lessons about
Ashley First Choirs Standing Work-in-progress preprint September 2020
airborne infection appeared not to have been learned by the time of the 2003 SARS outbreak when
this route was at first underestimated. Detailed studies developing the Wells-Riley concept were
lacking at that time and it is somewhat extraordinary that history appears to have repeated itself yet
again with the failure of the WHO to recognise airborne transmission at the outbreak of SARS-CoV-2
(Morawska and Milton, 2020).
Jiminez (2020) has been developing an Excel based model that can be used by anybody who is
prepared to work at understanding it and inputting the data. Herein lies perhaps the first problem
with this approach. The study by Caplin and Flick (2020) reveals difficulties in promoting the
engagement of choir directors with quantitative data. Data gathered from 23 of the “first choirs to
have stood” included location, age range, rehearsal space including floor area, distance between
singers and placing of singers and rehearsal pattern including breaks. Qualitative responses such as 2
hrs with 15 min break, Window and doors 2 metres in front, back and on the sides, in two rows were
unproblematic, as were simple measures such as floor area of rehearsal space in most cases, but the
request for “assessment of ventilation possibilities on a scale of 1 10” produced only seven
responses, not all of which gave a number between 1 and 10. The problem may have been the
formulation of the question. Data that would drive anything in the nature of a Wells-Riley based
spreadsheet need to be unproblematic for choir directors to obtain.
Assuming such difficulties can be overcome, there are then shortcomings with the Wells-Riley model
itself. Noakes and Sleigh caution that it assumes complete, homogenous air mixing. Recent studies
such as Shao et al (2020) have shown that this is a matter of no little significance. Jiminez is careful to
point out that his model is for aerosol transmission only and makes assumptions such as strict 2m
distancing. This difficulty might be overcome by the use of a dose-response model (Sze-To and Chao,
2009) but to the present author’s knowledge, no attempts to do this have yet been made. Jiminez
recognises many other shortcomings. The presence of just one adult with an underlying health
condition might confound the results which would surely need to be different for a youth choir as
opposed to a choral society. The model does not include such fundamental variables. It might need to
be run twice, once with an assumption of mask wearing and once without. The efficiency of different
kinds of mask and the ways in which they are worn has been shown by the parallel studies of Miller
(2020) and Srebric (2020) to be important variables.
Jiminez states that the model is deliberately kept simple so that it can be understood and changed
easily. This raises the difficulty that an impression of greater power will be attributed to the results
than might be justified, always an issue to be considered when parametric approaches are compared
with ordinal approaches. The actual power of the numbers is addressed by Jiminez who states that a
1% chance of infection obtained by the model in reality could be 0.2% or 5%. “It just won’t be .001%
or 100%” (Jiminez, 2020). Missing from this is any attempt to make a link between a probability of
between 0.2% and 5% and an L number. Does this level of probability count as “very unlikely” (1),
“unlikely” (2) or perhaps “likely” (4)?
A similar question arises in the case of the study by Shao et al which illustrates both the complexity
and power of computed risk approaches. Figure III below shows the modelling by these authors of a
classroom scenario in which design of ventilation is critical for reducing the risk of particle encounters.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Figure III
Classroom ventilation scenario in risk assessment through particle count
In this model, the students are largely passive whilst the teacher speaks for fifty minutes. In
accordance with work such as Asadi et al (2019), the projecting voice of the teacher emits significantly
more particles than does the passive breathing of the students. A badly placed ventilator at the rear
of the room results in a dangerous concentration of aerosols enveloping the student at the back of
the room. The risk for the students near the teacher at the front is less. This demonstrates amongst
other things that models of choir rehearsals in which a 6m spacing between conductor and front row
of singers are probably too simplistic. At the very least, a 6m spacing might give a false impression of
Shao et al make no attempt to create an interface between their model and the kind of data that a lay
person might be able to input. They evaluated risk as the total particle number passing through a
specified location during a period of simulation. This can be interpreted as the number of particles a
person could inhale during their time in a given location (𝐼risk). “Hot and safe zones were then
identified through five graduations of particle count from 1 - 104. From this, a control measure that
could be implemented by a lay person could be derived, though there is clearly much work to do in
providing information on the efficiencies and deficiencies of ventilation systems (the authors’ stated
There is still no solution to the problem of whether an (𝐼risk) of 103 or 104 counts as a category 4
(“likely”), a 5 (“very likely”) or perhaps only a category 3 “fairly likely”. It is clearly not possible is for
a lay person to make the (𝐼risk) calculation themselves, and similar calculations need to be made for
many necessary mitigations. This serves to illustrate just how complex the entire process is, and it
may be that the least unsatisfactory solution is simply to take prescribed control measures on trust.
As will be seen shortly, this is largely what the persons making decisions on behalf of the first choirs
standing did. By their own explicit admission, they could not be called “educated decision makers”.
Their decisions were made in all cases sincerely and carefully, but with the understanding of a scientific
lay person and in most cases without even the benefit of a simple control template such as that
provided by the Church of England. One possibility to be considered in this paper is that the lay risk
assessors in countries such as Norway, Germany or Australia were in fact over-cautious as a result.
The first choirs standing certainly exercised a high degree of precaution commensurate with
judgements of risk in situations of high uncertainty.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Method of the present empirical study
Given that there is known variability in infection rates across national boundaries, the first task in an
empirical study is to compile reference data on transmission, infection and death rates in different
nations. It was these data that largely informed the policies of national governments concerning
whether choirs could rehearse in their respective countries. Variables such as population density,
climate and median population age need to be considered also in making international comparisons.
Some order must also be brought to bear on timing. The data are therefore organised in reporting
periods as befits a method that is planned to be iterative. Two have initially been identified:
Period 1: A census day during the 2020 summer term
Period 2: A census day in September when most choirs return after a summer break
It is anticipated that cases may be subject to seasonal variation which may result in localised curbs on
choir activity. This is an unknown at the time of writing but will be considered in subsequent iterations
of the study. Initial data at national level will also need to be refined to bring out similarities and
differences at regional or even local levels across national boundaries.
Choir directors were invited first to submit a written account of how things appeared from their
viewpoint and what they were doing in practice. These accounts were then coded thematically and
analysed through the MAXQDA discourse analysis software. Through this process it was possible to
identify what the choir directors understood the risks to be and the control measures that they had
identified to manage them. Two surveys were then constructed in order to assess the relative
attention of each choir to the identified risks and measures. Table 1 shows the risks identified by the
choirs and the control measures they devised to mitigate them. Ranking is in order of the frequency
of mention, not the level of risk as quantified in the scientific literature. The table therefore reflects
perceptions that were held during the early stages of the pandemic. Greater detail is provided later
when case studies of each choir are presented.
Table 1
Perceived Risks and Controls Devised
Perceived Risk
Controls Devised
Contaminated surfaces
Enhanced cleaning
Removal of furniture/music stands
Close proximity of singers
Reduction of singer numbers
Spacing/distancing (various)
Floor markers
One-way systems
Sectional rehearsals
Infected singers attending
Voluntary quarantine
Poor ventilation
Air conditioning
Open doors and windows
Breaks to allow air to clear
Small rehearsal spaces
Larger rooms
Outdoor rehearsals
Older singers
Not members of the choir
Restricted attendance
Personal hygiene
Provision of sanitiser
Supervision of children
Long rehearsals
Shorter rehearsals
Cancellation of engagements
Ashley First Choirs Standing Work-in-progress preprint September 2020
The need for enhanced cleaning to mitigate the risk of contaminated surfaces appears to have
influenced the choirs the most, which may reflect the stress upon fomite transmission that occurred
early in the pandemic. Next most frequently mentioned were a variety of measures clearly influenced
by the messaging about social distancing that has been constant throughout the pandemic and
remains a key measure at the time of writing. The directors/managers were equally well-aware of the
dangers of airborne transmission, however. Given that the surveys took place before the WHO
changed its position on the issue of airborne transmission, the extent to which choirs perceived this
risk is notable. There was a strong recognition of the need to exclude singers with symptoms, but the
approach was almost invariably through voluntary, self-declaration measures. The risk from older
singers was recognised more by implication in that there is a sampling bias to choirs with younger
singers. Personal hygiene received fewer mentions that might have been expected, given the stress at
the time upon hand washing which continues at the time of writing. Possibly the choirs saw this as the
responsibility of individuals and assumed compliance. Only two choirs mentioned shortening
rehearsals, which is perhaps surprising given the likelihood that long rehearsals in enclosed spaces will
increase the airborne transmission risk. Finally, only one choir mentioned anything to do with travel,
and this was the cancellation of a tour programme. The risk of contracting and introducing infection
through car sharing or public transport on the way to rehearsal did not seem to be perceived or may
not have been relevant to the way particular choirs functioned.
Table 2 is derived from review of the scientific literature and identifies other risks that were not
mentioned by the choir directors.
Table 2
Risks for which choirs cannot devise control measures
Identified risk
Why a risk?
A high level of community infection rate (R
number above 1)
Increased risk of infected individuals attending,
including asymptomatic carriers
Transmission in the home
Close contact in the home reported as one of the
most potent means of transmission
Climate and weather
Viral potency and contagion likely to rise in
colder, wetter months
Cross-contamination from other activity
Risk proportionate to the diversity of activity,
perhaps more an issue for amateur than
professional singers
Higher transmission and infection reported
amongst ethnic minority communities and
The principal significance of Table 2 is that it is difficult or impossible for the choirs to devise control
measures to mitigate such risks. Unless the choir can move to another region or country, it has no
control over the risks arising from a high R number in the general population. There can be no control
over the climate and weather, other than perhaps planning concerts for the warmer summer months
when less contagion might be expected. With the possible exception of high-level professional groups,
choir directors are hardly in a position to demand changes to the domestic arrangements of their
singers or dictate arrangements for “social bubbles”. Similarly, if singers are involved in a number of
other activities, including going to work, playing sport, visiting family or eating in restaurants, the choir
conductor is hardly in a position to stop them. Ethnicity is a particularly difficult issue. Few conductors
Ashley First Choirs Standing Work-in-progress preprint September 2020
would want to risk or even consider a control measure such as barring certain ethnic groups, even if
it were done with the best of intentions on scientific evidence.
It is perhaps because the conductors were aware that they had little or no control over these risks that
they did not mention them. Nevertheless, they are significant risks. A choir might exercise the utmost
diligence regarding the matters over which it does have control, yet if the R number is high, the
chances of infected individuals attending rehearsals are proportionately high. If there are, for
example, unidentified weaknesses in the ventilation arrangements as suggested by Shao et al (op. cit.),
one individual could still infect many in the choir. For reasons such as this, choirs should perhaps
expect and accept that national governments and regional health authorities may impose lock-down
restrictions in response to trends in the populations for which they are responsible.
Case studies
Norway. Norway as a country is in a comparatively low risk category with a small, dispersed
population and reported death rate of only 4.4 persons per 100 000 (cumulative additional deaths by
6th June: 237). The resumption of choral singing was permitted from 22nd April onwards, with a guide
issued by the Norwegian Music Council and authorized by the Norwegian Health Council. A small,
qualitative study of choirs responding was undertaken at an early stage by Caplin and Flick (2020).
This identified twenty-three choirs that had recommenced singing from Norwegian choir week 19
onwards. Full details of the permitted conditions were given in Ashley (2020). The choirs ranged in
type from mixed youth choirs to female choirs and an adult male choir with age range 23 82 years.
The criterion for assessing the efficacy of the control measures was a request for a weekly update on
“proven infections”. Answers given were almost invariably “none” with just three instances of
Nidaros Cathedral, Trondheim
To be the first choir to stand is not to claim a prize or accolade but to bear the burden of risk on behalf
of the wider choral community. If any credit is due, it must go in this study the Nidaros Cathedral Boys
Choir in Trondheim. This may not have been the very first choir to stand, but it was the first to come
to the attention of the present author and participate in this study.
The conductor reported on the new rehearsal arrangements thus:
Week A: rehearsals within individual voice groups, which all are under 20 participants strong.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Week B: we split up the full choir into four sub-choirs of 20-22 participants, all SSAATTBB setup with
each of the SSAA voices consisting of at least one well experienced, one less experienced and one new
chorister. Those sub-choirs are giving us something we never have experienced before: new
groupings, more space between singers, new setups where choristers are unable to rely on “that one
sturdy soprano, who always sings correctly”.
Control measures implemented by Nidaros Cathedral Boys Choir
Time of year
Sunday Eucharist (broadcast) 10th May
Staged Performance for Norway Constitutional Day, 17th May
Approx 250 miles north of Oslo on south shore of Trondheim Fjord. 63oN 10oE.
Fourth largest city in Norway. Population density 539 /km2
Male cathedral choir in university city (portfolio-professional/education).
Large cathedral with natural ventilation. Doors kept open.
2.5 hour rehearsals reduced to 90 minutes
Age and
Normal complement. Approx. 60 boys aged 9-15 and 30 adult men
Reduced to 8 boys and 7 men for 10th May Eucharist
Maximum of 20 in any rehearsal. Increased use of sectional rehearsals with
reconfiguration to four equal sub-choirs. In each sub-choir, an experienced and
younger boy paired.
Proximity to
Normal rehearsal room not used. 300m2 space for exclusive use of choir in
adjacent building. Performers widely spaced in main area of cathedral (see
illustration below).
Antibacterial spaces at entries to room. Boys constantly supervised.
All music stands removed.
All travel and external engagements cancelled.
None reported
No reported infections.
Gosskor, Stockholm
The events leading up to the summer concert of the Stockholm Boys’ Choir show strong congruence
with the conditions reported in Sweden earlier, particularly voluntary restraint, schools remaining
open and a high death rate amongst the oldest members of the population. Many choirs in Stockholm
did voluntarily close down, but the boys choir continued to meet, preparing for its end of summer
term concert. This was streamed via a ticketing pay wall, and no live audience was present. The choir
director reported that Primary and secondary schools have been going on as usual here Our
biggest problem here has been the elderly and demented people (sic), so children´s activities are not
the biggest concerns for the authorities.” Notable in the table below is the degree of voluntary
restraint exercised by individual families, the director reporting that “the groups have naturally been
smaller because anyone with slightest symptoms have stayed at home.” On 6th June the reported
It has been difficult to assess cross-contamination (i.e. singers infecting their choirs as through contracting
the virus through another activity outside choir.) The assumption has been built in, therefore, that if the choir
is a professional one, or involves children under the control of their schools, the potential for cross-
contamination might be lower than in the case of amateur choirs where the membership is much more
Ashley First Choirs Standing Work-in-progress preprint September 2020
death rate was 43.2 persons per 100 000 (against 4.4 in Norway) with 4468 additional deaths having
accumulated by 6th June (Norway: 237).
Control measures implemented by the Stockholm Gosschor
Baltic Sea coast opposite Gulf of Finland. 59o N, 18o E.
Capital city. Population density 4800/km2
Secular Boys choir in in capital city (education/youth)
Floor to ceiling air conditioning system installed.
Rehearsal times shortened.
90 boys, 30 teenagers, 10 young adults, 10 mature adults.
50% absence in March, reducing to 25% absence in May
Normal rehearsal venue used (relatively small room of normal height)
Boys stand further apart, making use of absences. Some use of stickers and
Hand alcohol supplied. Room cleaning frequency increased.
Instructed not to attend if any symptoms.
Self-quarantine by parents reported as above.
No reported infections.
Of all the countries from which reports have been received, Germany is perhaps the least
homogenous. The 16 states have adopted their own approaches, often reacting to relatively localised
events. Federal directives have played less of a role although a national lock-down and curfew was
finally announced on March 22nd. Analysis of events before and after this date suggests that
Germany’s widely reported relative success in controlling the pandemic was achieved only through
much state-wide controversy and internal criticism. Differences in state approaches were notable in
data received from Berlin and Bavaria.
Staats und Domchor, Berlin.
The Staats und Domchoir of Berlin (state and cathedral choir) is part of the Universität der Künste
Berlin (arts university). Over 250 boys receive a musical education in groups ranging from Dominis
kindergarten to the Voces in Spe changing voice choir. The main concert choir consists of about 60
boys and 30 men. From this large complement a small ensemble of 5 boys and 5 men was selected to
prepare for a VE day memorial service on 8th May, with rehearsals resuming from 1st April. This is the
“first standing” performance considered here.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Control measures implemented by the Staats und Domchor, Berlin
April-May 8th
Berlin 121 miles inland from coastal town or Rostock. 53oN 13oE
Population density: 4000/km2
Large male choir associated with arts university, cathedral and federal state. 60
boys and 30 men in main choir. 250+ boys overall.
The largest indoor space in Berlin. Floor to ceiling air conditioners fitted.
Shortened rehearsal times with breaks for the air to clear.
Normal complement: 60 boys aged 10 14. 30 men
5 boys and 5 men with small group of period instruments
Large, circular cathedral. (“the largest indoor space in the city of Berlin”) All
rehearsals held in the cathedral itself or outside on steps. No rehearsal rooms
used. Singers spaced well apart (2-3m circle between each) in circular formation
(see illustrations below). Floor markers and one way system in use.
Enhanced cleaning regime. Superfluous furniture removed, singers retain own
Singers instructed not to attend if they have covid-like symptoms and asked not
to attend if they feel unwell.
It perhaps merits comment that the choir chose a circular formation, which is not recommended as
the singers will be facing each other. However, the voluminous space of the cathedral together with
the large diameter of the circle presumably acted as mitigations. It should also be mentioned that
the Staats und Domchor should not be confused with the adult Berlin Cathedral Cantory, as several
writers and press reports have done. An outbreak amongst members of the Cantory rehearsing for
Paul McCartney’s Liverpool Oratorio, did occur on 9th March.
Münchner Chorgemeinschaft, Munich
The choir had rehearsed twice after resuming on 15th June. Conditions in Bavaria clearly differed from
those is Berlin. Whereas the Berlin choir reported good support for the arts, Munich reported that
unless ticket revenues quickly returned to pre-covid levels the sector would be threatened. Across
Bavaria the choral sector was specifically targeted with instructions. During June, rehearsal was
permitted, but masks were required to be worn, singers to be spaced a minimum of 2m apart with 10
minute ventilation breaks for the room after every 20 minutes’ singing. Berlin, by contrast, had been
left more to their own devices, reporting that choral singing is subsumed within more general phased
returns to normality that they needed to interpreted for their sector.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Control measures implemented by the Münchner Chorgemeinschaft, Munich
Bavaria, 538 miles inland. 48o N, 12o W, elevation 524m
Not reported, though this was an adult amateur choir.
Fairly large room. Doors and windows left open.
15 mature adults, 32 elderly (over 20).
Only small, selectd groups attending. Parts rehearsed separately.
Singers asked not to attend if feeling unwell
Circle of 1 2m round each singer
Brisbane Cathedral (St. John’s)
At the time singing resumed, there had been only 102 deaths attributable to the SARS-CoV-2 virus in
the whole of Australia, a rate of 0.4 per 100 000 head of population, as opposed to 58.8 per 100 000
in the UK at that time. This is an almost meaningless statistic, given the size and average population
density of Australia. At the time Queensland (itself a large state, 667 900 miles2) had experienced no
new cases in the previous seven days with only two known positive cases remaining. Nevertheless,
the choir exercised a high degree of precaution, the conductor reporting that “we have no certainty
that we are doing the ‘right thing’ (and actually understand that there is no such thing) so the
responsibility weighs heavily on our shoulders.” Masks were attempted with the adult singers who
chose to abandon them. The boy choristers attend the Anglican Church Grammar School and stopped
singing at the same time as this school closed. The cathedral itself did not close, being considered a
“place of work”. Initially, five people were allowed in the building, so services were streamed by a
quartet, increasing to an octet drawn from the cathedral chamber choir when the permitted number
rose to 10. The boys first performed again on 25th June, recording a service to be broadcast via
Youtube on 28th June after rehearsing three mornings per week in school and once per week in the
cathedral (see illustration).
Ashley First Choirs Standing Work-in-progress preprint September 2020
Control measures implemented by choirs of St Johns Cathedral, Brisbane
Time of year
June/July (NB Winter in Australia. Mean temp 16o June, 15o July)
Brisbane -28o N, 153o E.
State capital of Queensland, Australia. Population density: 345/km2
Large cathedral in hot climate. Open doors and enhanced air circulation normal.
No additional arrangements practical.
Age and
10 boys (1)
6 young adults, 14 mature adults selected from chamber choir (2)
Proximity to
4 m2 space worked out for each singer. Clear separation (2m apart minimum,
see photograph)
Seats and stands are disinfected before and after every rehearsal.
Enhanced cleaning regime. Normal music desks and stands not used. Singers
keep own music copies.
Asked not to attend if feeling unwell.
New Zealand
Auckland Cathedral
Unlike Brisbane, the cathedral in Auckland, like all other churches, shut down completely early in the
pandemic (by 25th March). The government published early on a four-level alert system that was said
by the cathedral’s music director to be “very clear and effectively, though sensitively, enforced by the
police.” 25th March was the day that Level 4 alert (the highest) was enforced. Reduction to level 3
occurred on 27th April, level 2 on 13th May and level 1 on 8th. June. The cathedral reopened with six
singers and a congregation of 100 during level 2. Level 1 (8th June) allowed churches to return to
normal, including the use of choirs. Concerts had not been restarted by that date in Auckland although
the music director felt it would not be long. In spite of the unusually clear directions from the
government, however, he considered that he didn’t “really know what the contagion risks really are”,
though he felt that New Zealand hadn’t had had the large-scale confusion and inconsistency that
appears to have dogged the UK, Europe and the USA. For this reason, tabular data are not presented.
The choirs were able to resume normal activity very soon after the census date for this study. The
case makes for a particularly interesting contrast with Sweden, being perhaps the opposite end of a
continuum regarding how much individuals and individual organisations had to make their own risk
assessments and decisions.
Discussion and review: did the choirs get it right?
It must first be reiterated that the data presented above are a snapshot of conditions on the 6th June.
Since that date, much has happened. Many more scientific papers have been published, renewed
outbreaks have occurred (including in both Australia and New Zealand), governments have issued new
instructions. These changes will be reflected in the second iteration of this study which is intended to
take a snapshot of the situation in September 2020 when choirs resume from their summer breaks,
probably facing rising R numbers as temperatures fall and hard to predict regional resurgences of
contagion are reported if complacency sets in. That said, there are grounds to say that the choirs did
get many things right. All implemented control measures that were anticipatory of the advice that is
now being offered. None reported any infections. It would not be appropriate to claim that this alone
justifies any changes in policy or advice, but by the same token, should policy be driven by nothing
more than press reaction to the Amsterdam and Skagit “super-spreader” events?
Ashley First Choirs Standing Work-in-progress preprint September 2020
It is important to draw a conceptual distinction between precaution and risk. In general, a decrease
in precaution might be expected as scientific knowledge increases. High levels of precaution are
exercised when there is a high degree of scientific uncertainty. This has been seen to be the case for
most of the choirs. The Berlin choir, for example, exercised extreme precaution in reducing its
numbers from 60 boys and 30 men to 5 boys and 5 men, and by rehearsing either outdoors or in the
very large, ventilated space of the cathedral. Possibly they had been influenced by Berlin Cathedral
Cantory event.
The same is not necessarily true of risk. Estimations of individual hazards may rise or fall according to
the state of scientific knowledge. The most obvious example of this concerns aerosol transmission.
At the time the first choirs stood, the WHO were adhering to their scientific brief of 26th March which
claimed that “no airborne transmission had been reported”. This position was maintained until 9th July
against a considerable body of evidence that it was untenable. Lidia Morawska of the Queensland
University of Technology was the lead author of a critical position paper in the journal Environment
International (Morawska and Cao, 2020) and a more recent publication by Morawska and Milton
(2020) gained the support of 239 signatories to a letter. It is worth noting that history is simply
repeating itself here as the 2003 SARS outbreak also needed the invocation of alternatives to large
droplet explanations that were unable to account for the infection of individuals without sufficiently
close contact with known cases (Scales et al 2003). Four years later, Xie et al (2007) showed
retrospectively that airflow patterns within buildings indicated airborne infection.
One of the signatories to the Morawska document, a consultant virologist at the Leicester Royal
Infirmary, claimed on BBC Radio 4’s PM programme that the WHO had been selective in their
interpretation of evidence on order to maintain their own position (Tang, 2020). Allegations were
made, either that the WHO was obsessed by over-medicalised conservatism that discounts other
science, or that it feared stressing the importance of facial coverings because many developing
countries will be quite unable to provide them for the whole populations (see Royal Society/British
Academy, June, 2020). All this serves as a reminder that, though dispassionate objectivity is the
constant goal of science, it is not always achieved. What is of particular interest at this juncture is that
the first choirs standing were in some ways actually ahead of government advice in their precautionary
response to the risks posed by airborne infection. For example, the Berlin senate issued guidance on
the 12th August, three months after the Staats und Domchor had performed (see Table 3 below).
Two points stand out. The first is the emphasis on ventilation and air exchange, correctly anticipated
by the choir. The second is the requirement to wear face coverings, not anticipated or adopted by the
choir. The third is the absence from the senate guidance of several other measures taken by the choir
that would reduce the risk of singers standing in contaminated air (see above). There have been
several significant research projects that have released findings between the two dates. A much cited
but arguably overly simplistic experiment on airflow by Kahler and Hain (2020) produced somewhat
misleading results that perhaps set some choirs’ understanding of ventilation and flow backwards,
whilst the somewhat weightier study by Asadi et al (2020) on the relationship between phonation
intensity and aerosol production generated an angst right across the sector in most countries.
The high levels of angst are almost certainly attributable to the fact that the Asadi study did not
specifically address the issue of singing. It had been thought likely that aerosol production by singing
would lie towards the high end, perhaps on a par with or exceeding that of shouting or loud talking,
but in the absence of specific information, high levels of precaution predominated and certainly
influenced the judgements of Public Health England.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Table 3
Choir actions on 8th May compared with Senate instructions on 12th August
Berlin Senate 12th August
Staats und Domchor 8th May
The room must be ventilated regularly, ideally
by cross-ventilation
The largest indoor space in Berlin.
Rows of common singing must have an impulse
ventilation (ideally cross ventilation) of at
least 15 minutes
Floor to ceiling air conditioners fitted.
Continuous external ventilation (e.g. windows
on tilt or fully open) should be provided from the
beginning of the rehearsal or the event to the
All rehearsals held in the cathedral itself or
outside on steps
After the end of a rehearsal in which 60 minutes
of singing have taken place, the room must be
ventilated crosswise for 30 minutes, after which
the room must remain empty for two hours.
Before the start of the next rehearsal, again,
cross-ventilate for 30 minutes.
Rehearsal rooms not used.
Shortened rehearsal times with breaks for the
air to clear.
the minimum distance of 2 metres must be
maintained in all directions.
Singers spaced well apart (2-3m circle between
each) in circular formation.
A mouth-and-nose cover is required during
rehearsals and performances for singers and the
audience. However, it is strongly recommended
that singers and the audience wear the mouth
and nose protector for the entire duration of the
Face coverings not worn
Further complications arise from ongoing confusion between studies that address the behaviour of
aerosols (flow dynamics) and studies that address the production of aerosols. Echternach and
Kniesburges (2020) set up experiments specifically to identify the drift of aerosols in singing as
opposed to the projection of particles in simple air flow models such as that of Kahler. This study
addressed primarily the question of aerosol behaviour as opposed to production, though it gave some
indication of the number of particles that might be generated, confirming earlier speculation that
strongly emphasised consonants would project aerosols further. A significant finding concerned the
wearing of masks, concluding that these did significantly, though not completely mitigate the flow of
aerosols. The authors were reluctant, however, to recommend masks for professional singers. They
did show that the risk was significantly smaller towards the side of a singer than towards the front,
noting that the radial model of distancing (i.e. a 2m+ circle round each singer) might be revised. They
particularly stressed the need for a constant supply of fresh air to prevent aerosol accumulation
Mapping of infection risk on the probability scale through the Wells-Riley equation was carried out
during the second phase of the Colorado-Boulder study (Miller et al, 2020). This showed a significant
effect for the wearing of well-fitting masks, leading the authors to conclude that these should be
compulsory for a return to singing in schools. The study also placed a strong emphasis upon rehearsal
length and air change. Rehearsals should not exceed 30 minutes in length, and a time interval
corresponding to however long it takes to achieve a complete air change should intervene before the
next use of the room (Miller, 2020).
Ashley First Choirs Standing Work-in-progress preprint September 2020
The specific angst associated with the Asadi study has been addressed by Dirk Mürbe and colleagues
in Germany (Mürbe et al, 2020) and in the UK by the PERFORM project collaboration led by Pallav
Shah of Imperial College London, and popularly associated with co-investigator Declan Costello
(Gregson et al, 2020). This study demonstrated that at the quietest volumes, neither singing nor
speaking were significantly different to breathing, but at the loudest volumes (90-100dB), a significant
difference between singing and speaking could be observed. The authors stressed that the largest
changes they observed were nevertheless between the loudest and quietest volumes irrespective of
the type of vocalisation (greater than one order of magnitude). Differences between singing and
shouting, though statistically significant, were small enough for the authors to conclude that
guidelines could be produced in which volume, duration of vocalisation, number of participants and
the environment in which the activity occurs could be mitigations. No attempt was made to quantify
the risk or provide a means of assessing the efficacy of the suggested mitigations, but the data and
results were sufficient to convince PHE that there is no justification for treating singing as a uniquely
hazardous type of vocalisation. Against this must be considered that some English conductors appear
to have been excessively influenced by social media and press hyperbole, one rather brashly and
naively stating “it all depends on Declan Costello”. Neither should it be over-looked that the study
confirmed the possibility that any individual singer might be an unknown “super-spreader”. It remains
the case that it all depends on the ability of choirs to implement a range of mitigations that differ little
in principle from the mitigations necessary for any activity in which there is collective indoor
Finally, another study that is awaiting peer review has been undertaken for the Association of
Professional Musicians in Sweden. Noting the work that had been undertaken in other countries, this
study recommended the following controls for singing in Sweden:
joint singing in closed up rooms should not exist
If rehearsals are happening, the volume of the room must be as big as possible, with very good
ventilation possibilities. Air must be replaced, not recirculated.
The duration of a rehearsal must not exceed 15 minutes, and the empty room must be properly aired
Unventilated rehearsal rooms should be shut down. Ventilated rooms could possibly be used, given a
few hours between users.
(SYMF, 2020)
Whilst clearly broadly in agreement with the emerging consensus that all indoor gatherings in which
voices are used (for whatever purpose) must be time limited and require large, well ventilated spaces,
a rather higher degree of precaution is recommended than was exercised by the Stockholm Boys
Choir. Sweden seems, by this criterion, currently to be travelling in the opposite direction to the UK.
For all the painstaking work that has been carried out in large studies such as the Colorado-Boulder
collaboration, the question of how to express and quantify risk remains. James Weaver, Director of
Performing Arts and Sports for the US National Federation of High Schools stated that
We are not using a live virus, we are not infecting participants, we are not allowing participants to be
infected while doing labratory experiments so something that we can’t answer is, you know, if the play,
will they be infected, we can’t tell you that (Weaver, 2020).
Arguably, the first choirs standing did use a live virus in that they took the risk that the live virus would
not be present in their spaces in sufficient quantity to cause an infection but it could have been.
Ashley First Choirs Standing Work-in-progress preprint September 2020
They were able, as a result, to say that if they sing, they will not be infected. That, fundamentally, is
the point of an empirical study and it is the basis of familiar statistical risks such as that of death in an
air accident. Of course, no ethics committee would sanction the deliberate construction of a study
that would intentionally expose participants, including children, to a live virus. Nevertheless, it was
the judgement of the choir directors and managers that the risk of this was acceptably low that kept
the children acceptably safe. It was this judgement that counted, not a clinical, hermetic exclusion of
the virus and it has been clear that an extremely high degree of precaution was exercised. Judgement
was expressed in intention to act on a set of control measures informed by whatever understanding
was available at the time. Whilst the present author felt obliged to uphold the position that choral
singing was “high risk”, these choirs in their various locations and circumstance took the position that
choral singing with stringent controls was “acceptable risk”, even for children. It is the stories of these
choirs, not a number, that gives substance to the meaning of “acceptable risk”.
The experimental studies that have been published since the first choirs stood have moved in a clear
direction. Provided that controls relating to limitations on numbers of singers, time of rehearsals, and
above all ventilation and air change are implemented, the level of precaution has indeed reduced in
proportion to the increase in scientific knowledge. The fact that the first choirs standing anticipated
most of these controls might be taken as a demonstration that some trust might be placed in choir
directors to manage risk. This does not eliminate risk, however, but it is empirical study, not
experiment, that will remain the final arbiter as is the case with transport safety. As choral singing
resumes in the UK, an empirical study is by default taking place. We all hope that the result will be
“no infections reported” from which the conclusion “if they sing, they won’t be infected” can be
Andersen, K., Rambaut, A., Holmes, E. and Garry, R. (2020) The proximal origin of SARS-CoV-2, Nature
Medicine, 26: 450-455.
Asadi, S., Wexler, A., Cappa, C., Barreda, S., Bouvier, N. and Ristenpart, W. (2019) Aerosol emission
and superemission during human speech increase with voice loudness, Nature, 9:2348.
Asadi, S., Wexler, A., Cappa, C., Barreda, S., Bouvier, N. and Ristenpart, W. (2020). Effect of voicing
and articulation manner on aerosol particle emission during human speech. PloS one, 15(1), e0227699.
Ashley, M. (2020) Where have all the singers gone, and when will they return? Prospects for Choral
Singing after the SARS-CoV-2 Pandemic. ABCD Choral Directions Research,
Borch DZ, Sundberg J. Some Phonatory and Resonatory Characteristics of the Rock, Pop, Soul, and
Swedish Dance Band Styles of Singing. J Voice. 2011;25(5):532-537. doi:10.1016/j.jvoice.2010.07.014
Caplin, T. and Flick, M. (2020) Database of choirs returning to singing in Norway. Used with permission.
Chin, A. et al (2020) Stability of SARS-CoV-2 in different environmental conditions, The Lancet Microbe
Delve Initiative, the, (2020) Balancing the Risks of Pupils Returning to Schools. DELVE Report no. 4, 24th
Ashley First Choirs Standing Work-in-progress preprint September 2020
van Doremalen, N., Bushmaker, T., Morris, D., Holbrook, M., Gamble, A., Williamson, B., Tamin, A.,
Harcourt, J., Thornburg, N., Gerber, S., Lloyd-Smith, J., de Wit, E., and Munster, V. (2020) Aerosol and
Surface Stability of SARS-CoV-2 as Compared with SARS-CoV-1, New England Journal of Medicine,
Correspondence, April 16th.
Echternach, M., Gannter, S. et al (2020) Impulse dispersion of aerosols during singing and speaking,
medRxiv preprint doi:
Ferguson, N., Laydon, D. et al (2020) Report 9: Impact of non-pharmaceutical interventions (NPIs) to
reduce COVID-19 mortality and healthcare demand. London: Imperial College Covid-19 Response
Team. DOI:
Folkehelseinstituttet (2020) Kunnskap, situasjon, prognose, risiko og respons i Norge etter uke 18.
Norwegian Public Health Authority.
Goldman, E. (2020) Exaggerated risk of transmission of covid-19 by fomites, The Lancet Infectious
Diseases, 20: 892-893.
Gregson; Watson; Orton; Haddrell; McCarthy; Finnie; et al. (2020): Comparing the Respirable Aerosol
Concentrations and Particle Size Distributions Generated by Singing, Speaking and Breathing.
ChemRxiv. Preprint.
Havers, F., Reed, C. et al (2020) Seroprevalence of Antibodies to SARS-CoV-2 in 10 Sites in the United
States, JAMA Internal Medicine. doi:https//10.1001/jamainternmed.2020.4130,
Heald-Sargent, T., Muller, W, Zheng, X et al (2020) Age-related differences in nasopharyngeal severe
acute respiratory syndrome coronavirus 2 (SARS-CoV-2) levels in patients with mild to moderate
coronavirus disease 2019 (COVID-19). JAMA Pediatrics online.
Heneghan, C., Brassey, J. and Jefferson, T. (2020) SARS-CoV-2 viral load and the severity of COVID-19,
University of Oxford, Centre for Evidence Based Medicine, 26th March.
Jiminez, J. (2020) Introduction to the Covid-19 aerosol transmission estimator, University of Colorado-
Boulder online recorded webinar, 22nd July.
Kähler, C. and Hain, R. (2020a) Flow analyses to validate SARS-CoV-2 protective masks About distance
rules, mouth-nose protection, particle filtering respiratory protection, filter materials and mask
manufacturing, Preprint from University of the Bundeswehr, Munich, 11th April.
Kähler, C. and Hain, R. (2020b) Making music during the pandemic - what does science advise? About
infection risks when singing choirs and making music with wind instruments. Preprint from University
of the Bundeswehr, Munich, 8th May.
Ashley First Choirs Standing Work-in-progress preprint September 2020
King, A., Cosdtello, A., Friston, K., Khunti, K., McKee, M., Michie, S., Pagel, C., Haque, Z., Pillay, D.,
Pittard, A., Pollock, A., and Scalley, G. (2020) COVID-19: what are the options for the UK?
Recommendations for government based on an open and transparent examination of the scientific
evidence (The independent SAGE report). London: The Independent Scientific Advisory Group for
Konnai R, Scherer RC, Peplinski A, Ryan K. Whisper and Phonation: Aerodynamic Comparisons Across
Adduction and Loudness. J Voice Off J Voice Found. 2017;31(6):773.e11-773.e20. doi:10.1016/j.jvoice.
Lipsitch, M. and Viboud, C. (2009) Influenza Seasonality: Lifting the fog, Proceedings of the National
Academy of Sciences of the United States of America (PNAS), 106(10): 3645-3646.
Miller, S. (2020) Performing Arts Aerosol Study, Round one preliminary results, University of Colorado-
Miller, S., Nazaroff, W. et al (2020) Transmission of SARS-CoV-2 by inhalation of respiratory aerosol in
the Skagit Valley Chorale superspreading event, medRxiv preprint submitted to Indoor Air, 15th June.
Ministry of Health, Government of New Zealand (2020) Aotearoa/New Zealand’s COVID-19
elimination strategy: an overview.
Moldofsky, H., and Patcai, J. (2011). Chronic widespread musculoskeletal pain, fatigue, depression and
disordered sleep in chronic post-SARS syndrome; a case-controlled study. BMC neurology, 11: 37.
Morawska, and Cao, J. (2020) Airborne transmission of SARS-CoV-2: The world should face the reality,
Environment International, June, 139: 105730.
Morawska, L. and Milton, D. (2020) It is Time to Address Airborne Transmission of COVID-19, Clinical
Infectious Diseases, July 6th ciaa939. doi: 10.1093/cid/ciaa93.
Murbe, D., Fleischer, M. et al (2020) Aerosol emission is increased in professional singing, Preprint
from Berlin: Charité Universitätsmedizin, Department of Audiology and Phoniatrics.
Nardell, E. ((2016) Wells Revisited: infectious particles. Quanta of mycrobacterium tuberculosis
infection don’t get them confused, Mycrobacterial Diseases, 6: 231-234. Doi:10.4172/2161-
Nardell, E. (2016) Wells Revisited: infectious particles vs. quanta of mycrobacterium tuberculosis
infection don’t get them confused, Mycrobacterial Diseases, 6: 231-234. Doi:10.4172/2161-
Naunheim et al (2020) Safer Singing During the SARS-CoV-2 Pandemic: What We Know and What We
Don’t. Journal of Voice,
Ashley First Choirs Standing Work-in-progress preprint September 2020
Noakes, C. and Sleigh, P. (2008) Applying the Wells-Riley equation to the risk of airborne infection in
hospital environments: The importance of stochastic and proximity effects Indoor Air Conference,
17-22 August 2008, Copenhagen, Denmark - Paper ID: 42
Park, Y., Choe, Y., Park, O. Park, S, Kim, Y., Kim, J. et al (2020) Contact tracing during coronavirus disease
outbreak, South Korea, Emerging Infectious Diseases.
Public Health England (2019) UK measles and rubella elimination strategy 2019. London: PHE
Ratnesar-Shumate, S. et al. Simulated sunlight rapidly inactivates SARS-CoV-2 on surfaces. J. Infect.
Dis. (2020) doi:10.1093/infdis/jiaa274/5841129.
Riley, E., Murphy, G. and Riley, R. (1978) Airborne spread of measles in a suburban elementary school,
American Journal of Epidemiology, 107: 421432.
Scales, D. et al. (2003) Illness in intensive care staff after brief exposure to severe acute respiratory
syndrome, Emerging Infectious Diseases, 9 (10): 1205-1210.
Somsen, GA, et al. (27 May 2020). Small droplet aerosols in poorly ventilated spaces and SARS-CoV-2
transmission. The Lancet Respiratory Medicine (published online).
Spahn, C. and Richter, B. (2020) Risk assessment of a coronavirus infection in the field of music,
preprint fourth update, July 17th. Freiberg: Hochschule fur Musik.
Stage, H., Shingleton, J. et al (2020) Shut and re-open: the role of schools in the spread of COVID-19 in
Europe, MedRxiv preprint,
Kamerlin, S. and Kasson, P. (2020) Managing COVID-19 spread with voluntary public-health measures:
Sweden as a case study for pandemic control. Clinical Infectious Diseases.
Shao, S. et al (2020) Risk assessment of airborne transmission of COVID-19 by asymptomatic
individuals under different practical settings, arXiv preprint, Cornell University
Stage, H., Shingleton, J., Ghosh, S., Scarabel, FD., Pellis, L. and Finnie, T. (2020) Shut and re-open: the
role of schools in the spread of COVID-19 in Europe, medRxiv preprint.
Srebric, J. (2020) Performing Arts Aerosol Study, Round one preliminary results, University of Maryland.
Ashley First Choirs Standing Work-in-progress preprint September 2020
Stein-Zamir, C. et al (2020) A large COVID-19 outbreak in a high school 10 days after schools’
reopening, Israel, Euro Surveill, 25(29):pii=2001352.
Sveriges Yresmusikerforbundsymf (2020) [Swedish Union of Professional Musicians.
Tang, J. (2020) Today Programme, BBC Radio 4, 6th July.
Royal Society/British Academy (2020) Face masks and coverings for the general public: Behavioural
knowledge, effectiveness of cloth coverings and public messaging, Preprint rapid review report, 26th
Sze To, G. and Chao, C. (2009) Review and comparison between the WellsRiley and dose-response
approaches to risk assessment of infectious respiratory diseases, Indoor Air, 20: 2-16.
Weaver, J. (2020) Performing Arts Aerosol Study Preliminary Results 2, online webinar, 6th August.
Williams, R. (2020) Opera singer’s six-month coronavirus nightmare, Caerphilly Observer, Wednesday
12th August.
Xie, X., Li, Y. et al (2007) How far droplets can move in indoor environments revisiting the Wells
evaporationfalling curve, Indoor Air, 17: 211-255. doi:10.1111/j.1600-0668.2006.00469.x
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
We investigate the effect of school closure and subsequent reopening on the transmission of COVID-19, by considering Denmark, Norway, Sweden, and German states as case studies. By comparing the growth rates in daily hospitalisations or confirmed cases under different interventions, we provide evidence that the effect of school closure is visible as a reduction in the growth rate approximately 9 days after implementation. Limited school attendance, such as older students sitting exams or the partial return of younger year groups, does not appear to significantly affect community transmission. A large-scale reopening of schools while controlling or suppressing the epidemic appears feasible in countries such as Denmark or Norway, where community transmission is generally low. However, school reopening can contribute to significant increases in the growth rate in countries like Germany, where community transmission is relatively high. Our findings underscore the need for a cautious evaluation of reopening strategies that ensure low classroom occupancy and a solid infrastructure to quickly identify and isolate new infections.
Full-text available
The global impact of COVID-19 has been profound, and the public health threat it represents is the most serious seen in a respiratory virus since the 1918 H1N1 influenza pandemic. Here we present the results of epidemiological modelling which has informed policymaking in the UK and other countries in recent weeks. In the absence of a COVID-19 vaccine, we assess the potential role of a number of public health measures-so-called non-pharmaceutical interventions (NPIs)-aimed at reducing contact rates in the population and thereby reducing transmission of the virus. In the results presented here, we apply a previously published microsimulation model to two countries: the UK (Great Britain specifically) and the US. We conclude that the effectiveness of any one intervention in isolation is likely to be limited, requiring multiple interventions to be combined to have a substantial impact on transmission. Two fundamental strategies are possible: (a) mitigation, which focuses on slowing but not necessarily stopping epidemic spread-reducing peak healthcare demand while protecting those most at risk of severe disease from infection, and (b) suppression, which aims to reverse epidemic growth, reducing case numbers to low levels and maintaining that situation indefinitely. Each policy has major challenges. We find that that optimal mitigation policies (combining home isolation of suspect cases, home quarantine of those living in the same household as suspect cases, and social distancing of the elderly and others at most risk of severe disease) might reduce peak healthcare demand by 2/3 and deaths by half. However, the resulting mitigated epidemic would still likely result in hundreds of thousands of deaths and health systems (most notably intensive care units) being overwhelmed many times over. For countries able to achieve it, this leaves suppression as the preferred policy option. We show that in the UK and US context, suppression will minimally require a combination of social distancing of the entire population, home isolation of cases and household quarantine of their family members. This may need to be supplemented by school and university closures, though it should be recognised that such closures may have negative impacts on health systems due to increased absenteeism. The major challenge of suppression is that this type of intensive intervention package-or something equivalently effective at reducing transmission-will need to be maintained until a vaccine becomes available (potentially 18 months or more)-given that we predict that transmission will quickly rebound if interventions are relaxed. We show that intermittent social distancing-triggered by trends in disease surveillance-may allow interventions to be relaxed temporarily in relative short time windows, but measures will need to be reintroduced if or when case numbers rebound. Last, while experience in China and now South Korea show that suppression is possible in the short term, it remains to be seen whether it is possible long-term, and whether the social and economic costs of the interventions adopted thus far can be reduced.
Full-text available
Hand washing and maintaining social distance are the main measures recommended by the World Health Organization (WHO) to avoid contracting COVID-19. Unfortunately, these measured do not prevent infection by inhalation of small droplets exhaled by an infected person that can travel distance of meters or tens of meters in the air and carry their viral content. Science explains the mechanisms of such transport and there is evidence that this is a significant route of infection in indoor environments. Despite this, no countries or authorities consider airborne spread of COVID-19 in their regulations to prevent infections transmission indoors. It is therefore extremely important, that the national authorities acknowledge the reality that the virus spreads through air, and recommend that adequate control measures be implemented to prevent further spread of the SARS-CoV-2 virus, in particularly removal of the virus-laden droplets from indoor air by ventilation.
The lack of quantitative risk assessment of airborne transmission of COVID-19 under practical settings leads to large uncertainties and inconsistencies in our preventive measures. Combining in situ measurements and computational fluid dynamics simulations, we quantify the exhaled particles from normal respiratory behaviors and their transport under elevator, small 15 classroom, and supermarket settings to evaluate the risk of inhaling potentially virus-containing particles. Our results show that the design of ventilation is critical for reducing the risk of particle encounters. Inappropriate design can significantly limit the efficiency of particle removal, create local hot spots with orders of magnitude higher risks, and enhance particle deposition causing surface contamination. Additionally, our measurements reveal the presence of 20 a substantial fraction of faceted particles from normal breathing and its strong correlation with breathing depth.
On 13 March 2020, Israel's government declared closure of all schools. Schools fully reopened on 17 May 2020. Ten days later, a major outbreak of coronavirus disease (COVID-19) occurred in a high school. The first case was registered on 26 May, the second on 27 May. They were not epidemiologically linked. Testing of the complete school community revealed 153 students (attack rate: 13.2%) and 25 staff members (attack rate: 16.6%) who were COVID-19 positive.
Group singing events have been linked to several outbreaks of infection during the CoVID-19 pandemic, leading to singing activities being banned in many areas across the globe. This link between singing and infection rates supports the possibility that aerosols are partly responsible for person-to-person infection. In contrast to droplets, the smaller aerosol particles do not fall to the ground within a short distance after being expelled by e.g. a singer. Aerosol particles hover and spread via convection in the environmental air. According to the super-spreading theory, choir singing and loud talking (theater and presentations) during rehearsals or performances may constitute a high risk of infectious virus transmission to large numbers of people. Thus, it is essential to define the safety distances between singers in super-spreading situations. The aim of this study is to investigate the impulse dispersion of aerosols during singing and speaking in comparison to breathing and coughing. Ten professional singers (5 males and 5 females) of the Bavarian Radio Chorus performed 9 tasks including singing a phrase of Beethovens 9th symphony, to the original German text. The inhaled air volume was marked with small aerosol particles produced via a commercial e-cigarette. The expelled aerosol cloud was recorded with three high definition TV cameras from different perspectives. Afterwards, the dimensions and dynamics of the aerosol cloud was measured by segmenting the video footage at every time point. While the median expansion was below 1m, the aerosol cloud was expelled up to 1.4m in the singing direction for individual subjects. Consonants produced larger distances of aerosol expulsion than vowels. The dispersion in the lateral and vertical dimension was less pronounced than the forward direction. After completion of each task, the cloud continued to distribute in the air increasing its dimensions. Consequently, we propose increasing the current recommendations of many governmental councils for choirs or singing at religious services from 1.5m to the front and 1m to the side to a distance between choir singers of 2m to the front and 1.5m to the sides.
Background: The COVID-19 pandemic has spread globally, causing extensive illness and mortality. In advance of effective antiviral therapies, countries have applied different public-health strategies to control spread and manage healthcare need. Sweden has taken a unique approach of not implementing strict closures, instead urging personal responsibility. We analyze the results of this and other potential strategies for pandemic control in Sweden. Methods: We implemented individual-based modeling of COVID-19 spread in Sweden using population, employment, and household data. Epidemiological parameters for COVID-19 were validated on a limited date range; where substantial uncertainties remained, multiple parameters were tested. The effects of different public-health strategies were tested over a 160-day period, analyzed for their effects on ICU demand and death rate, and compared to Swedish data for April 2020. Results: Swedish mortality rates fall intermediate between European countries that quickly imposed stringent public-health controls and countries that acted later. Models most closely reproducing reported mortality data suggest large portions of the population voluntarily self-isolate. Swedish ICU utilization rates remained lower than predicted, but a large fraction of deaths occurred in non-ICU patients. This suggests that patient prognosis was considered in ICU admission, reducing healthcare load at a cost of decreased survival in patients not admitted. Conclusions: The Swedish COVID-19 strategy has thus far yielded a striking result: mild mandates overlaid with voluntary measures can achieve results highly similar to late-onset stringent mandates. However, this policy causes more healthcare demand and mortality than early stringent control and depends on continued public will.