RGC sessions are designed to be a discovery learning process. While the facilitator will have some idea of what to expect after preparing for the workshop, s/he will be simultaneously learning the language from the participants and guiding them through the process of checking and elicitng data. 

The linguist must be comfortable with a great deal of ambiguity and uncertainty. It is challenging to consistently choose helpful activities for studying and eliciting data, while not really knowing what will be discovered along the way. The facilitator also must be willing to test grammar with native speakers in the full knowledge that some analyses will be found to be incorrect. Experienced RGC facilitators note that there are always surprises each day as participants confirm and reject various insights. 

There is little need to train participants on general linguistics topics (such as phonetics and typological linguistics). The only need is to bring awareness to the features of the participant's native language. For this reason, RGC does not include any lectures, only brief explanations of how to do each new activity and its purpose. This allows more of the workshop time to be spent observing language patterns and constructions, then deciding the writing system based on what is discovered.

In all the activities, there needs to be a balance between doing careful linguistic analysis and spending too much time of a particular word grammatical detail. Go slow enough that the majority of participants are hearing the correct sounds, understanding some functions of the morphemes, observing the morphophonological patterns, tentatively agreeing on spelling, and writing correct grammar in the notebooks. However, keep things moving. 

Occasionally, participants will need time to discuss issues that arise, such as whether two adjacent roots should be joined as words, whether or not a certain noun has a plural form or any number of other issues. Allow each person to have a say in the matter, but don’t allow such discussions to go on more than about 5 minutes, unless the discussion is important for the activity. At an appropriate point, bring the participants back to the goal of the activity, and encourage discussions on the issue(s) to continue during the next break time. Often the issue can be more easily adressed in a later session when all relevant grammar is collected and discussed. 

For a practical writing system, it is more important to find the sound changes of morphemes than to fully understand all the functions of morphemes. Keep to the grammar that will help you arrive at the goal of agreeing on an informed writing system. Often a broad view of the grammar, such as investigating most morphemes in most environments using representative data, will be a better foundation for making orthography decisions than a deep but narrow view of the language, such as learning all there is to know about a few morphemes in the language. 

In a language with numerous sound alternations at morpheme boundaries, stifle your linguistic curiosity to exhaust all the syntactic functions of a particular morpheme. Remember that your primary goal of the workshop is help the language group arrive at a tentative working orthography and keep the research focused on that goal. 

No matter how well prepared you are for the workshop, there will be one or more sessions that do not go as well as expected. In most cases, the cause is either that the data is not as expected or that the participants did not understand the aim of the activity. 

When data is not as expected:

When the data is not as you thought it would be based on the texts or what you know of related languages, try to still make the session useful, such as by revising a frame until it can be used to collect the intended data or by finding out enough details of the grammar to know how to collect the data in another way. There is no shame in not having all the answers. Remember that participants are the ones who are the experts on the language, not you. Your job is to find a way for them to communicate information about the language. 

If you suspect that difficulties in facilitating a session are causing the participants to loose their enthusiasm for the work, stop and change to another activity. Later, rethink the possible ways of getting the data you need for the writing system. If possible, pull a participant aside before the next workshop session and verify that your new activity or frame will successfully elicit the intended data. If you are stuggling to collect data related to particular verb forms, collect one or more full paradigms of complicated verbs that can be used as models for paradigms of other verbs.

When participants do not understand the goal of the activity:

When the participants are not responding as you expect, it may be because they do not understand the aim of the activity or what you expect of them. You may need to...

• Give a brief explanation for why the activity is important.

For example, you can say, "The goal of this activity is to have a list of correctly spelled words in the dictionary that all language developers can refer to."

• Explain exactly what you would like the participants to do.

For example, you can say, "Slowly read this list of words while everyone else listens for any vowels that are different than the rest." 

• Explain what you mean by asking questions, since questions are sometimes used as rebukes in certain cultures.

You can say, “Sometimes when I ask a question, I already know the answer. I ask the question because I want to know if you also know the answer. Then I will know that we are together in the work.” 

The linguist’s responsibility is to ensure that the participants have adequately understood all the issues involved in making decisions related to the writing system. This includes readability, transferability to the national language, similarities with related languages, etc. While the actual decisions should be left to the participants, the linguist can have a profound influence on these decisions, which should not be abused or taken lightly. 

RGC activities are designed to display language data in a systematic way, which empowers partipants to make thoughtful, consistent choices. The linguist plays an important role in helping organize the data and identifying (potentially) problematic writing decisions. Rather than cititng qualifications or experience, a linguist’s best influence is calling attention to the language data. 

As the workshop progresses, the linguist should also keep a rough estimate of the number of minimal pairs or grammatical distinctions made only by tone, vowel quality or other phonemic distinctions. If there are a significant number of these not distinguished in the writing,* the facilitator must a) discuss with participants how the reading could negatively be affected and b) propose some ways of representing the distinctions in writing. 

*See page "End of Phase 1" for further details on predicting readability.

Often when a language group makes a writing decision that a linguist experienced in orthography design might consider unwise, the decision will not so dramatically effect reading or writing that it would be a waste of time to try to use the writing system in spite of that decision. But there are cases, where fluent reading may be extremely difficult until changes are made. For example, a language with with a high functional load of tone or [ATR] quality may need to add more tone markings or vowel symbols to effectively distinguish meaning. In such cases, the linguist may need to recommend that the project not be supported until a writing system for improved reading is found. The changes do not necessarily need to be those suggested by the linguist, but they do need to result in reading fluency. 

After a tentative writing system is agreed-upon in Phase 1, the facilitator will need to provide ongoing training and support in using the new system. Writing in one's own language is a high-level skill that takes years to develop. Even a well-chosen writing system by a language community can be worth little if they do not learn to consistently use that writing system. Inconsistent writing hinders reading fluency and comprehension, devalues the reading materials, and produces inconsistent writers as readers see what appears to be an acceptable example to follow.

In a 2018 study, RGC pioneer Tim Stirtz found that most workshop participants either forgot their writing system, did not yet understand how to use it, or for other reasons did not consistently apply it when developing written materials. Among the 9 language communities that were studied, there was an average of 1 inconsistency* for every 10 words (Stirtz 2018:6-7). The most common types of inconsistencies involved similar vowels, tone representation, word breaks, sound changes at morpheme boundaries, or similar consonants. The potential for negative effects from this inconsistency was significant. However, with 2 months of intentional training (which took place over several years), these inconsistencies were reduced by an average of 70%! By contrast, a tenth community that did not receive this training had little, if any, improvement in writing consistency. 

The training given to the 9 language teams merely involved an editor checking each word of each book in context, and for each inconsistency found, reminding the authors of their agreed-upon writing system. Through this process, there were occasional changes made to the initial writing decisions when the community was in agreement, and then all materials were revised accordingly. The initial writing decisions, changes to writing decisions, as well as all revisions to books in application of the writing decisions, were all at the discretion of the community members. Although the editor gave feedback and training, the communities had the final say in all language decisions.

So while agreeing upon a writing system is not enough for writing consisency, ongoing training after the workshop is proven to be highly successful. See Stirtz's paper (2018) Linguistic checks: The missing link in modern language development for more information on facilitator checking.  

*For the purposes of RGC, "writing inconsistencies" refers to those that are unintentional, differing from the written standard previously agreed upon by the community for a particular language, and occurring in draft materials later to be published, such as for teaching reading or for use in schools or churches.