Inclusive Event Guidelines
So, you’ve got yourself an initial plan. Great! It’s time for some speakers.
There are typically three ways that invited speakers could be incorporated in your programme. Most academic conferences will have the majority of their speakers selected via a competitive anonymous review of papers/abstracts, with perhaps one or two keynote addresses from invited speakers. Some conferences and other events may include panel talks from three to five invited speakers on a certain cohesive topic. Or your event could involve only invited speakers. These formats have different strengths for different contexts, but in the first instance, we would caution against fully invited speaker lists for at least the context of a general academic conference. This is because without some measure of anonymous review it is harder to reduce the impact of implicit bias on finalised programmes. But some types of events (e.g., career advice days, project workshops etc.) may require arrangement by purely invited speakers.
However they are incorporated into your programme, invited talks are important parts of your event when considering inclusivity. Often they are seen to confer status on the individuals giving the sessions, as they afford certain speakers an amount of increased focus and prestige within a schedule. For example, keynote addresses hold a central structuring and framing role, and are usually ran with no parallel sessions and a longer time slot than other sessions. Moreover, speaker selections also tend to track perceived seniority (or ‘big name’ status) in the profession, and as a result tend to be the least representative places at an event. This is unfortunate, as one factor in repelling under-represented groups from the philosophy profession is the inability of some to ‘see themselves’ in that role; the absence of marginalised persons from these higher status speaking slots contribute to the professorial stereotype remaining one of whiteness and maleness in particular.
On this note, we recommend that wherever possible you consider speakers from under-represented groups first, give more importance to their commitment to your event, and have a significant portion from under-represented groups on your list of potential speakers. To reiterate information from §2.1, you may have considered allowing for some flexibility in the final date of your event, with invited speakers having some say in this decision. The most important voices to have in this conversation will be the speakers from marginalised groups that have been invited to give sessions. These measures together make it more likely that your keynote speakers will be more diverse. We make some specific recommendations below.
These recommendations may be challenging, particularly the first time you attempt to reorient your organising priorities in this way. But this is important work to ensure fairer treatment of current researchers from marginalised groups and to encourage early career researchers from these groups in their academic careers. And you are not alone: there are resources out there that can help you source speakers with relevant specialisms and experience to invite.
Your first port of call will usually be both your own acquaintance with researchers in the relevant area, and that of your colleagues (whether local or international). This can be leveraged to help diversify your list of potential candidates. Ask a wide range of colleagues their opinion on potential keynote speakers. This will involve getting the opinions of people from PhD students upwards, and from all backgrounds, which will help to expand your pool of invitees.
The UP Directory - http://www.theupdirectory.com/
This internet-based resource provides a list of academics from departments across the world who self-identify as members of marginalised groups. Although fairly US-centric, it’s highly useful in establishing contact with marginalised academics.
Diversity Reading List - https://diversityreadinglist.org/
Although its primary focus is on providing resources for diversifying teaching curricula, the Diversity Reading List can also be used to get an idea of who works in various specialisms within philosophy. The core of the website is a searchable database of papers written by philosophers from under-represented groups, listed by specialism, keywords and author. The website’s listings also often include a link to the staff or personal webpage of the person who wrote each paper, so the contact details for potential speakers should be fairly simple to find.
Speakers that have rejected your invitation
It’s almost inevitable that some potential speakers will reject your offer of a keynote address. You can, however, use this to your advantage in some cases with a simple request of further recommendations. Former invitees are perfectly placed to know other potential speakers with relevant experience for your event, and may be able to recommend more junior colleagues that lack immediate name recognition. This request can either be made pre-emptively in the initial email (e.g., “if you are unable to accept this invite, would you be able to recommend any other researchers with the relevant expertise to give this keynote address?”) or once they have provided their response. Please ensure that you do so in a way that avoids making your correspondent feel tokenised, by emphasising that you are looking for other talented researchers whose work is as relevant and high quality as their own. Most researchers will be willing to offer such recommendations, but please note that they are not obligated to, and that this should not replace your sourcing of potential speakers by other means.
Multiple invited slots
There are many scenarios in which you may have multiple invited speaker slots to fill. You may have more than one keynote address, or you may have one or more panel talks. Alternatively, your event could be invite-only! In these circumstances, there are some recommendations that we have regarding your overall inclusivity.
Adopt a hard policy of securing no less than 40% of invited speakers as those from marginalised groups, with an underlying aim of achieving equal representation
Adopt a policy whereby no less than 40% of invited speakers for any particular event element (e.g., keynotes, panels) are given by people from marginalised groups
Commit to a policy of never going ahead with events (or elements therein, e.g., panels) that involve all white male participants
There are very few circumstances where this is unavoidable, and where you believe that you are subject to these circumstances, we urge you to return to the resources detailed above
If you commit to these recommendations, you are well on your way to a more inclusive event! However, please note that these recommendations do not deal with guidelines on choosing respondents (if this is an element of your event), which is considered below.
Calls for papers/abstracts
It’s important to give as much information as possible in your call for abstracts or papers, and advertise as soon as is practicable given our other recommendations. Alongside obvious information (such as relevant dates), we recommend that this should include (where possible):
Inclusion of demographic details (e.g., self-identifications of membership of any marginalised group) on the submitted cover page(s), voluntarily, based on whether those submitting feel comfortable disclosing this information
Accessibility information for the main venue(s)
Details for the named contact for personalised accommodations/accessibility requests, and/or the process for making a request
Commitment to inclusivity related to the event and in the proposed method for review
Commitment to accommodating the additional needs of all speakers selected, with this information to be collected at the registration stage or communicated in advance via the named contact
Commitment to extending the deadline (up to a specific period) for those that need extra time for medical and/or accessibility reasons
Any standardisation features for submissions (font sizes etc.)
Any proposed method(s) of digital conferencing to be used
If you are following these guidelines, we would also appreciate an acknowledgement that you are committed to these recommendations
These features together will allow you to plan more readily (and earlier) for any additional accessibility measures, and for ensuring diversity in your speaker selection. On the latter point, more detail is given in our section on the review process on how demographic information could be used responsibly without undermining blind review. The inclusion of information regarding the main venue(s) is crucial for inclusivity at your event. For example, those with mobility issues, visual or hearing impairment, or fluctuating health conditions may decide that the risk of submitting to your event is too high, since the cost of having to withdraw after committing to a slot is large. Note that this is also good for you too – you’ll reduce the need of speakers to withdraw closer to the event.
On the standardisation of submission formatting, the purpose here is to help eliminate more areas of implicit bias in the review process. It is easy to have a negative first impression of a submission based on the font type, size and other aspects of general formatting, because each feature has some connotations attached (whether positive or negative) when used in a variety of contexts. To take an extreme example, you may think that the use of the font Comic Sans would be inappropriate for a paper submission to a major conference. This may be because of the general impression that it is light-hearted, implies a lack of serious content, and not suitable for professional communications. You may think that this case is obvious, but there are always going to be more subtle borderline cases, where some may feel a font type is professional but others feel it isn’t. You may think that this has little impact on your decision-making, but even a minimal impact will be cumulative over many events. And it is also the case that those from working class backgrounds may not be as fluent in the art of presenting information in the way you find most ‘professional’ or ‘appropriate’, as it is a part of cultural and institutional knowledge that is not often communicated outside of (relatively) privileged circles. Having some standardisation in place should thus allow a more inclusive process.
To provide some additional guidance, we’ve drafted an example CFP and CFA for your use in Appendix A and Appendix B.
CFP vs. CFA
There are costs and benefits of both the choice to advertise for papers and the choice to advertise for abstracts. CFPs allow for the presentation of more detailed and finished material, and often an increased availability of the highly drafted material for consideration by attendees before and after your event. This comes at a cost of a heightened barrier of entry, which may mean that some will not be in a position to commit to your CFP when they otherwise might have with a CFA. This is especially the case where you have not provided enough detail regarding the accessibility of your venue and/or your funding situation regarding childcare/speaker expenses at this stage, thus discouraging those with mobility issues and junior colleagues (which are least likely to have access to independent funds for travel etc.). CFAs allow for a greater variety of applications and of potential speakers, partially by lowering the cost of entry in terms of time investment. However, it is therefore likely that you will have much more material to consider in general.
Because of the potentially emancipatory effects of adopting a CFA (over a CFP), we suggest that you give serious consideration to this option, even in cases where you are fairly set on advertising via CFP.
Speaker selection (inc. review)
Once you’ve received your prospective abstracts/papers, you should have a set of anonymised submissions, and a corresponding set of cover sheets containing information (including demographic information) regarding each potential speaker. So far, so good. But how should we decide between them?
There’s two features to consider in your review process: anonymization, and how to ensure adequate representation at your event. These considerations may appear in tension, and it is true that there may be some trade-offs when you design your review process. Some things you may be considering are: allocating a certain proportion of speaker slots to those self-identifying as from a marginalised group (we’d recommend a 50% allocation for this option); allocating a certain proportion of speaker slots to early career and/or postgraduate researchers; or allowing self-identifications of marginalisation as a method of breaking ‘ties’ between submissions for inclusion in the schedule. This list is not exhaustive and each option is suitable for different contexts. If you have other options for balancing these considerations, please let us know for future editions of these guidelines.
For now, we’re going to run through a full review process that you might consider, along with the reason for each step, to give an example of how this could be resolved.
Step 1.One member of the committee takes the submissions, converts them to an accessible format (e.g., HTML, RTF, accessible PDFs), and renames them in an anonymous format (e.g., Person A, Person B, etc.); this is repeated for the corresponding cover sheets
This step foregrounds the need to provide files that are accessible across different operating systems (Windows, MacOS, Linux etc.) and incorporates easily usable measures for anonymization
Step 2.The anonymiser then sends the edited submissions (not cover sheets) to all of the committee to read and consider
By centralising full information to one person throughout, this minimises the risk of implicit bias affecting speaker selection
Step 3.Each member of the committee reads these submissions and assigns a rough ranking to each (from first to n’st)
This allows basic data that can be used to produce average rankings; these can be used throughout the process, mitigating the effects of potentially unintended deference to the opinions of more senior and/or outspoken colleagues in speaker selection
Step 4.A meeting to decide on the final schedule is convened, with committee members bringing their individual rankings and the anonymiser bringing the data from corresponding cover sheets
Note that this can be achieved via online methods, e.g., email if needed, but that individual rankings should be made by all members before sharing
Step 5.Rankings are collected and a mean average (to the nearest decimal) is taken of the rankings given to each submission, providing a new consensus ranking; the top number of submissions (i.e., those with rankings nearest to 1) corresponding to the number of speakers you need is your draft schedule
The creation of an average ranking gives you a clear way to consider submissions comparatively, whilst maintaining higher levels of objectivity in the subsequent speaker selection (again, mitigating the undue influence of any particular committee member)
Step 6.Discussion of each submission can proceed at this stage, to see if this changes the overall order and/or opens up opportunities for reconsideration in conjunction with demographic data
At this stage, the shape of the conference (e.g., diversity of topics) should be considered, and debate about the relative merits of your submissions can happen, with a view to having a second draft schedule
Step 7.The anonymiser then discloses demographic information for the submissions that have this attached
This allows the use of demographic data to ensure diversity of speakers without undermining the process of anonymous review
Step 8.Discussion of the effect of demographic information on the overall selection of speakers is considered by the committee; if you’ve already got a good range of demographics, good job; if not, consider swapping out some of the lower ranked submissions with those in effective ties (e.g., those in the region of up to one integer lower than the lowest in your draft schedule) but from marginalised persons
You can use this opportunity however you see fit, but we urge you to seriously consider alterations to your draft schedule where your overall speaker allocation falls below recommendations set out below
Step 9.You now have a final selection of speakers for your event!
Note that this can easily be adapted for quota models by moving the role of demographic data to initial demarcations of submissions under consideration for each respective speaker allocation
Note: Please ensure to avoid researching your speakers based on their internet presence throughout this process, as this may bring in elements of implicit bias
This process may appear needlessly complex at first glance, but the cost here is worth it. First, it doesn’t add a significant amount of extra work to committee members, by sticking to the general processes of anonymous reviewing. Second, the effect it can have on the inclusivity of your event(s) can be transformative. Even if there are some events which still have disproportionately few researchers from marginalised groups as speakers, the effect on the macro-scale is much larger: more events will be more inclusive over time. And this is a huge benefit for the events themselves and our discipline as a whole. It is also important to note that this process is still fully anonymous and controlled as far as possible for the effect of bias throughout.
Parallel to our section on invited speakers, we recommend the following commitments to guide your selection process.
Adopt a policy of securing no less than 40% of speakers as those from marginalised groups, with an underlying aim of achieving equal representation
Commit to a policy of never going ahead with events (or elements therein, e.g., panels) that involve all white male participants
Due to the current demographic make-up of the discipline, these recommendations may be difficult to fully implement. The level of control you have over this will obviously depend on who submits to your event, and the quality of those submissions. But it is important to have a set of over-arching objectives to guide your actions, and subsequently increase the representation of marginalised peoples at events (even if this is a gradual process).
Scheduling (inc. parallel sessions)
Your next step will be developing your full schedule based on the speakers you have selected during review. This will often have the aim of maintaining a spread of specialisms and topics across the event. We’d also recommend that speakers from marginalised groups are spread throughout the event, to avoid risking the tokenisation of a certain session or sessions. There are some exceptions to this. For example, at larger cross-specialism (and other) events, you may seek to have a breakout session in which researchers from marginalised backgrounds can network and support each other, and this may be lead/chaired by exclusively marginalised speakers. It is important to remember that providing this space is not a replacement for increased representation in the main and invited speaker schedules.
A quick note on parallel sessions. If you are pursuing this structure, you will need to consider the impact of this on respective attendance to sessions and how this might disadvantage marginalised peoples. This will need to be handled on an individual basis, but we have one recommendation in this area: attempt to ensure an adequate audience for each session, and importantly, plan the attendance of your committee around your estimates. Of course, you will, by necessity, have at least one committee member in each session, but any surplus members can be allocated towards talks that you feel may lose out in direct competition. This may also need to be changed on the day, based on actual audience distribution (and this is covered in more detail in §2.4). Granted, this may mean that some committee members do not see all the talks they would like to, but it ensures you maintain your responsibility to maximally support the cohesion of your event, its value to all participants, and the well-being of all speakers.
You may be considering having respondents to some or all speakers, in order to have a detailed comment on each talk which engages with the work on a deeper level than typical in a Q&A. If you are doing so, we’d recommend three groups that would be suitable to contact for this purpose: the pool of people whose submissions were not selected for a substantive session, your local postgraduate community (or communities, where this is a cross-institutional event), and your local staff base(s). We would recommend that this opportunity is offered to early career researchers, postgraduates and researchers from marginalised groups in the first instance. As with your schedule of speakers, you should then spread your respondents across the event as far as possible, ensuring that you avoid the tokenisation of some session(s) by concentrating all and only your respondents (and speakers) from marginalised groups therein.