Despite the tremendous potential of the DDP method for collecting a huge corpus of words in a minority language in a very short period of time, it has not been integrated into our organization's planning. In this workshop we will look at:

• Case studies of several teams who have done the DDP process and it's results
• How to organize a word collection workshop

• Sharing and discussion of current methods for word collection for dictionary building
  • Toolbox interlinearization adds new lexical entries
  • Teams begin with an African wordlist to fill out
  • No organized method. User enters words as he/she encounters them
• How successful are these methods?
• How much time does it take to arrive at a useful lexicon or wordlist?

Dictionary Development Program

Saafi Saafi?

Present and discuss

Read and discuss

Read and discuss

• Data is already tagged according to semantic domain.
  • If entered into FLEx, SIL has plans to allow translators to search semantic domains for words and propose translations.
• Vast amount of data in a short period of time
• If definitions are also supplied, the wordlist can be posted to an internet repository focusing on language preservation
• Funds can be obtained from from outside organizations to help manage the data with the goal of language preservation.

• Problems due to dialect differences
• Cost of leading the workshop?
• Others?

How was the data handled in the various case studies and what might have been done differently?

If funding were available for word collection workshops, how might this be handled?

The purpose of the workshop is to collect words to put into a dictionary. You may also want to have the participants give a simple one-word gloss or short definition in the national language so that you have some indication of what the words mean. Since you will be using semantic domains to collect words, the words will be automatically classified by domain. So the result of the workshop is a (glossed) classified word list. If possible you should try to type the words during the workshop. However some people have found this impossible due to logistical problems, such as a lack of electricity or an absence of typists. Ideally you would come out of the workshop with a massive computerized database of around 20,000 words. At the least you should have around 10,000 words on paper. You will have something similar to, but better than, a thesaurus.

This workshop can be a lot of fun. Motivation can be very high. People like their language and exploring it can be fascinating. Moving through the pathways of the mental lexical net can be a lot of fun. Word games such as Scrabble, Password, and Outburst are based on the process of thinking of words and associating one word with another. Your enthusiasm, especially at the beginning, will set the tone for the entire workshop. I highly recommend that you use the DDP word collection method in a workshop setting. Some people have tried to have just one or two people collect the words. However, not only does this take much longer, but the results have always been disappointing. Someone who is exceptionally good at word games and can think and write (or type) quickly could conceivably do the job by himself in a month or two. If a person spent ten minutes on a domain it would take 30 days to finish. Realistically it would take much longer. Experience has shown that the synergy of working in a group results in a far greater number of words being collected. It helps to have one person writing down the words while the others think. In most societies people enjoy working together. One person’s thoughts stimulate another’s. The social interaction helps maintain interest in what could otherwise become boring and tedious. So I recommend that you recruit 20-30 people and organize them into small groups of 3-4 people. This gives enough small groups to work through all the domains in a two-week workshop. There is no theoretical limit to the number of people involved. However logistics may limit the number that can be handled. Having a variety of people means that you will be working with their collective mental lexicons. People have various areas of expertise. Some people have a larger vocabulary than others. Some people are better than others at pulling words out of their minds. People who have been educated outside of the language area may have better analytical skills, but often their ability in their mother tongue is more limited than someone who has spent their whole life speaking the same language. Older people will know different words than younger people. For all these reasons it has proven to be far better to have a group of people work together. Another major reason to hold a workshop is that the job gets done fast. Producing a dictionary will take many years unless ways are found to speed up the work.

You will need to recruit twelve to twenty participants. There is quite a bit of work involved in processing all the domains. Theoretically there is no limit to the number of participants. However there should be twelve to twenty participants if you want to finish within two weeks and do a thorough job. The number can vary depending on the quality, education, and experience of the participants. Recruiting a variety of people will mean that people with areas of expertise and experience are available to work on specialized domains. It is absolutely crucial that the participants be of good quality. Although it is hard to predict who will be good at this task, several qualities are essential. They need to be able to write their language. It is possible to include a few illiterates, but they will have to work with someone who can write. They need to be able understand the domain templates. If their English is not sufficient, someone else will have to translate the template for them. This will significantly slow down the process. The domain templates have been (or are being) translated into French, Spanish, Portuguese, Swahili, Chinese, Malay, Indonesian, Thai, Nepali, Tamil, Hindi, and Amharic. It is hoped that they can be translated into other major languages as well. The most important quality in participants is the ability to think well. A few people never seem to catch on to the idea of semantic domains or what is required. This kind of person can seriously detract from the work, slowing others down, including words that don’t belong to the domain, arguing without understanding what they are talking about, and generally causing everyone else problems. Make sure the people you invite have the ability to comprehend abstract concepts and understand instructions. When recruiting people, it is important that they understand what they will be doing. Explain to them the purpose and methodology. If you are going to pay them, or at least pay their expenses, make sure they clearly understand what is being offered. More than one workshop has been nearly wrecked because the participants got the impression that they would be paid a lot. If they are being asked to volunteer their time, make sure they clearly understand this. Consider working with a few people ahead of time to train them in the procedure. This may be especially helpful if you already have a small team you have worked with and whose abilities you are confident of. It takes a while for people to catch on to the procedure and become productive. I found that the participants were thinking of twice as many words toward the end of the workshop as they were the first week. Having a few people at the beginning, who understand the procedure, will help the others catch on sooner. If you do not speak the language, it is very important to have someone who can explain the procedure in the language of the participants. One task that can be done ahead of time is to translate the domain labels. This will enable a few people to get a good grasp of the entire system and what a domain consists of. If you do this, you can type it up and make a copy for everyone. However the participants should be encouraged to think of a good label for a domain after they have finished thinking of the words that belong to the domain. You will need to obtain books of pictures for several of the domains. It is necessary to have pictures with scientific names for the domains for plants and animals if such are available. It is insufficient to have words in the dictionary defined as ‘a species of animal’. Getting a book of pictures will eliminate this problem and greatly improve your dictionary. Today it is possible to find guides to the flora and fauna of every area of the world. The biology department of a university or the national museum of your country should be happy to help. You should be able to find good books for the mammals, birds, fish, reptiles, and insects. Good books on the smaller animals may be harder to find. There may be dictionaries of plant names available for the national language. Pictures of other cultural features, such as houses, cultural artifacts, or ceremonies, may also be useful. You can also search the internet for pictures of plants and animals. The semantic domain templates need to be printed. It will take several hours to print all 1792 domains (471 pages), depending on the speed of your printer. Be sure you have sufficient paper and don’t overheat your printer. The domain templates have been prepared for printing and are included in the document Templates.doc. They will print in continuous format with a page break after each secondary heading. If someone else has held a word collection workshop, you may be able to borrow their printout. Make sure it is the most recent version. (This is version 5.) Put each secondary subsection (there are about 60, e.g. 1.1, 1.2, 1.3, etc.) in a separate folder and write the section number and heading on the outside of the folder. It is good to staple the sections together to keep them in the correct sequence. This will help to divide the task into smaller achievable tasks. Since the task is so big, finishing a folder will help the participants feel a sense of accomplishment and see progress. I keep track of progress on the blackboard (total number of folders, total number of domains—1792, number of folders completed, number of domains completed, number of words typed). I have the participants write their names on the folder when they are finished, so that I know who worked on it. The participants will be writing the words on separate pieces of paper. You will need to purchase several reams of lined paper. If you don’t use lined paper, things can get messy and hard for the typists to interpret. Have the participants use a separate page for each domain and write the domain number at the top. You will need to check that the participants have not skipped any domains. The participants will need lots of workspace. There will be lots of discussion, and hence lots of noise. If everyone is together in one small room, the noise can get distracting. It may be helpful to separate the group into different rooms to cut down on the noise level. Some of the participants preferred sitting outside under trees. In many societies it is preferable to be together. They will be writing all the time and will need tables on which they can spread out several pieces of paper at a time (the domain folder, the domain they are working on, related domains, extra paper, etc.). Since they will be working together on a single piece of paper, they will need a long enough table so that more than one person can sit side by side so that both can read the paper together. Each small group should have its own table. However some participants preferred sitting on the ground and writing on their laps.

• 12-20 participants.
• A printout of the domain templates.
• 60 folders, one for each major section of the domains. Staple each section together and put it in a folder. Write the number and domain label on the outside of the folder. Put all the folders in a box.
• 3000 pieces of lined paper for the participants to write the words on.
• A copy of the list of domains for each participant.
• A copy of the instructions for each participant.
• Pens for each of the participants, one blue and one red for each participant.
• Two computers and two typists.
• A copy of the blank dictionary loaded into Toolbox on the computers, so that the typists can type the words.
• Stapler and staples.

Give each of the participants a copy of the accompanying document, “Instructions for filling out the domain templates” (ddp5_instr_for_participants.doc). Use it as an outline for giving them instructions. Give a general overview of what the workshop is all about. Explain the basic steps in producing a dictionary. Explain how this workshop fits in. Give everyone a copy of the sample templates (ddp5_sample_templates.doc). This document contains several of the domain templates, which you can use for practice. Read through them. This will give everyone a feel for what they are and enable you to explain specific aspects of a domain template. If you would prefer practicing with other domains, you can pull others out of the list, or simply photocopy several pages of the templates. I print a full page with some extra domains to give the context, but I would only have the participants fill out one domain on the page. For instance on the second page we would only do ‘Head’. I’ve picked some sample domains that give the participants a variety of semantics and part of speech. The samples I had in mind were ‘Wind’ ‘Head’ ‘Lose consciousness’ ‘Save from trouble’ ‘Supernatural being’ ‘Tend a fire’ ‘Sleep’ ‘Walk’ ‘Cut’ ‘Quick’ ‘Possible’. I would pick at least half a dozen of these to do together. Pick a couple that consist mainly of nouns, a couple that consist of verbs, and a couple that consist of abstracts. Each participant should have a copy or be able to share a copy. Fill out the sample domains as a group. The goal is to get all the participants very familiar with the procedure and what kind of words we are looking for. Don’t skimp on the training. If they don’t get the idea, you will end up with lots of junk. The brighter participants will feel like they understand after doing just two or three of these. But the slower ones need to watch for quite a while before they catch on. You should monitor these practice sessions closely and point out any problems. Try to catch the mistakes early. If they aren’t catching on, do some more practice. Expect to do a full day of training. It won’t be wasted. You should continue the training until the participants have caught on to the idea of using their mental network to jump from word to word. Give everyone a copy of the list of domains (ddp5_list_of_domains.doc) and have everyone review all the domains. Everyone needs to be familiar with all the domains, even if they never work on them all. This is because there is a tendency to “expand” a domain beyond what was intended and include words from other domains. If the domain labels have been translated ahead of time, give each participant a copy and have them read it through. They can refer to it throughout the workshop whenever they are unsure if a word might belong somewhere else. Explain the concept of a semantic domain. Explain that the words of their language are in their heads. Our job is to get those words out and onto paper. Explain that each person organizes the words of his language in his mind. He forms links between words and organizes them in groups. When we talk about a topic, the mind uses these links to quickly bring to mind the words we need to say what we want. This process is subconscious, but we can use the links to think quickly of all the words that are linked together. Draw a spider web on the blackboard as an illustration. Each word is connected to other words. Show how ‘sleep’ is related to ‘wake, nap, bed, bedtime’ and many other words. Define ‘semantic domain’ as a central idea and all the words directly related to it, or as an area in the mind that contains a number of words with similar meaning, or as a topic that we talk about and all the words we use to talk about that topic, or as a father and his children. You may need to find some metaphors that will effectively communicate to them the nature of the mental lexicon. It is very important to explain the mental process that we are using to collect words. They are not just translating a list of English words. They are employing the mental network of lexical relations to rapidly jump from word to word. If we allowed our minds to freely roam the mental pathways of lexical relations, we would be jumping from one domain to another in rapid succession. This might result in a lot of words, but it would be hard to systematically cover all areas of the lexicon without leaving huge holes. So we limit the process to a single domain at a time and work through all the domains of the language. But we still want to encourage the participants to explore their mental pathways. If they get outside of the narrow confines of a domain, have them get a blank piece of paper and write down the words they are thinking of. They should then locate the other domain on the list of domains and put the correct number at the top of the page. The words can be added to the proper domain later. However you should encourage them to try to stick to a single domain. Otherwise they will duplicate the effort of some other group working on the other domain. The important thing is that they learn to use the powerful mental abilities of creativity and free association. Experience has shown that the biggest problem using this method is that the participants merely translate the sample words. Many participants never grasp the method and never use their mental network. They never think whether the words they are translating belong together in the domain. They never evaluate if the DDP domain matches one of their own. The workshop turns into a translation exercise. Therefore it is very important to train them sufficiently. The domain templates are merely a guide and should be used merely to suggest possibilities. Their job is to use the associations in their heads to think of words that belong together in their language. The elicitation questions can help them to ensure that they have done a thorough job of thinking through their domain. But they should never be used blindly. Explain the organization of the list. Point out the main divisions and subdivisions. Explain the hierarchical organization, that each group of domains is related to the major heading. Explain the concept of a ‘citation form’ for the words in a dictionary. Explain the citation form for nouns, verbs, and adjectives. Other parts of speech may need special handling or may not be inflected. The rules will vary for each language. If you have not worked out the rules for citation forms previously, do so now as a group. If a problem arises during the workshop with unusual words, have the participants ask you for help. In English the only parts of speech that are problems are nouns and verbs, because they are the only ones that can be inflected. The rules for English are:

• For nouns give the singular, not the plural.
• For verbs give the bare root (or stem), not an inflected form (e.g. run, not runs, running, or ran).

Explain the use of idioms, phrases, and slang expressions. Define and illustrate each from the English examples and from their own language. Warn them not to literally translate English idioms. Encourage them to think of natural idioms, phrases, and slang expressions in their language. Organize the participants into small groups. It may be good to work out a rotation system and schedule. This may be important if you have some people that aren’t very good. It would be too bad for a capable person to get stuck with a partner who made things difficult for him for the whole two weeks. If the participants know each other, they may prefer to work with their friends. Let each small group do one or two domains. Then have the whole group review each domain, making corrections and additions. This will help you catch problems. Explain what to do with specialized domains. It is best to have people work on domains that they are familiar with. Clergy should work on Religion. Policemen and government officials should work on Government and Law. Doctors should work on Health. Encourage the participants to choose domains that they are most familiar with or at least sound interesting to them. Explain that they will have to keep moving in order to finish. They should only spend about fifteen minutes on a domain, although some domains will take less and some more. They should not spend a lot of time trying to think of one or two more words. When it becomes difficult to think of other words, move on. There is a principle that governs memory and recall. When a person is trying to remember a list of things, his rate of recall gets slower and slower as he approaches the end. He may recall the first items in a very short period of time, but take a very long time to remember the last few. So in order to make good progress, the participants will have to learn to quit work on a domain when it gets hard to think of more words. A little healthy competition is not bad. Explain to them that the first time this procedure was tried, fifteen speakers of Lugwere in Uganda collected 10,000 words in ten days. Challenge them to beat that record. It shouldn’t be hard. As of early 2006 this workshop has been held for approximately 200 languages. In the second workshop to use the method, 30 speakers of Lunyole collected over 14,000 words in ten days. The world record now stands at 23,000 words in two weeks. However some languages are only collecting around 8,000. It is not clear why some groups do so much better than others. There are a number of factors, including the enthusiasm of the group, the education level, and how well the semantic domains fit their culture. We are still learning.

When the participants have finished a folder, have them give it to you. Check it over to make sure the pages are in order, the words are in the correct citation form, and the words all belong to the domain. If the participants did not do a thorough job and have obviously missed many words, consider giving it back to them or breaking up that particular group. I have found that great diplomacy is needed in dealing with participants who were not able to do a satisfactory job. It may be possible to find one of the extra tasks listed below that they can do and put them to work on it. It is important that the participants review all the domains in a section before they start filling out the first one. The first domain in a section is often very general. If the participants have not reviewed the entire section, they might put specific words that belong in a later domain under the first more general domain. If at all possible, work out the citation forms ahead of time. There is an accompanying document (Citation.doc), which explains the principles of choosing a citation form. Give the participants enough practice to become familiar with the correct form. I have found that some participants gave an imperative form since the English examples often sound like imperatives. This is especially a problem when an elicitation question asks for something like steps involved in a process. The participants may give imperative forms in the typical style for giving instructions, ”(First) peel, (then) cut….” Or they may be explaining, “We peel, we cut….” Some domains tend to elicit plural forms. For instance in the domain for Food, participants may be completing the statement, “We eat .” The answer is often in the form of a list of plurals, “Apples, pears, grapes, oranges, lemons….” I sometimes failed to check their work, and later found entire sections had been filled out with the wrong citation forms. This will result in a lot of duplicate entries and cleanup work later on. Don’t worry too much about spelling. If the language does not have a dictionary, chances are the people aren’t very literate. A dictionary can be a powerful tool in analyzing the phonology of a language, developing an orthography, and standardizing the spelling of words. The dictionary will help to pinpoint problems. The goal in this workshop is to collect the words. A trained person can go through the dictionary later and fix any problems with the spelling. In many cultures people like to work in groups, rather than working alone. So one of your jobs will be to organize the participants into small groups. You can let them team up with their friends or you can assign them to a group based on your knowledge of their skills. You will need one person in each group who can read the materials. You will also need one person who can write down the words. It is essential that there be one person in the group who understands the process well. It is also essential that there be at least one person who is good at thinking of words. You may get more work done if you have them work individually. However I have never run a workshop in which anyone worked alone. The participants always wanted to work in groups. The reasons are probably because they were not confident working alone and most people enjoy working together. The synergy of working in a group seemed crucial for this process. You will have to decide how big the groups should be. Working in large groups means that you have fewer groups and may not get as much done. However a large group may be more business-like, keep things moving, and produce more words per domain. Working in pairs means that you have more groups working on more domains at one time. However a small group may not think of many words, and if the members are not especially good, the domain may be filled out wrong. Some groups tend to waste time in unnecessary discussion. You will have to find the best organization for your workshop. The best strategy is to monitor the work being done and solve problems as they arise. Some people are better at this task than others. I’ve found that about half the participants caught onto the idea and were very productive. The rest were unproductive and made lots of mistakes if they worked on their own. Some merely translated the English examples and were unable to go beyond them. We tried to partner someone who was good with someone who was having trouble grasping the concepts and procedure. That way we didn’t get domains that were filled out wrong and had to be redone. However burdening your best people with someone who is slow may merely drag the good people down. It may be best to rotate partners. If someone is not very good, or is hard to work with, one person doesn’t get stuck with them the whole time. Another option is to put the good people together and let them be productive. Put the slower people together in one group and work with them yourself. During the first few days closely monitor the work of each group and look for problems. It is extremely important to catch problems early on. Messy work will result in a huge cleanup job for someone later on. Strive for a balance between freedom and creativity on the one hand, and careful, accurate work on the other. As much as possible it is good to have a number of different specialties represented in the workshop. If none are available during the workshop, it may be necessary to find them later in order to get specialized vocabulary. If the participants make good progress and finish early, you might consider letting them look for specialists who could contribute specialized words in their field. Each culture has unique areas of specialty. Eskimos build igloos; Ifugaos build rice terraces; Maguidanaons cultivate coconuts. It may be necessary to set up extra domains if one of these areas of specialty contains many words. Every effort has been made to provide a place for everything, but it is impossible to predict all specialized cultural domains. You should help the participants set up the domain and number it under the appropriate major domain. In one workshop the participants asked where they should put the words having to do with gathering and eating white ants, a local delicacy. I set up a special domain and the participants listed over 30 specialized lexical items belonging to this domain. Please notify me of each such case, so that future editions of the list can be modified as needed. Section 9 ‘Grammar and discourse’ is the domain for grammatical words, functors, and affixes, and will require special attention. It may be best for a linguist to work with a few of the better participants on it, perhaps outside of the workshop so that the better participants can be available to help the others during the workshop. It may take more time to fill out section nine than the other sections, because it is often necessary to write out entire sentences and not just words. I found that after the workshop had been going a couple of days, the participants knew what they were doing and needed very little help from me. At that point I picked the best two participants and put them to work on Section 9. I spent a fair amount of time explaining what was needed in each domain, but normally left them to work alone. It was sometimes necessary to directly elicit the words in the most difficult domains. Generally they were able to complete a domain after a brief explanation from me. I was very pleased with the huge number of functors they collected and the excellent example sentences they wrote out. Don’t skip section 9 just because it looks difficult. Keep the participants moving. You don’t want to be a harsh taskmaster, but there are a lot of domains. It takes about fifteen minutes to do a good job on a domain. Some domains may take more time, some less. If you have six groups working eight hours a day, it will take approximately eight days to finish. My experience is that the first and last days of a workshop are usually unproductive. So you may be hard pressed to finish all the domains in a two-week workshop. If you have lots of participants and if they work hard, they may finish early. It may be helpful to have them review the domains they have finished and try to add more works. My experience is that the participants added an average of four words per domain the second time through. Have them work on domains that they have not worked on before. You may want to have them work on the domains that were done early in the workshop, since they usually improve over time. If they add words to a domain that has already been typed it is critical that they write in different color ink so that the typists know what has been typed and what hasn’t. This is why I recommend getting both blue and red pens for each participant.

The first time I held one of these workshops, I had the participants gloss all the words in English. This seemed to slow them down. I thought it might be better if they worked as much as possible in the vernacular, so their minds could better access their mental lexicon. So in the second workshop I only had them write the vernacular words. The words were typed during the workshop, and afterwards I had someone enter a simple definition for each entry in the database. Unfortunately he sometimes didn’t know what a word meant. Sometimes words were unfamiliar to him because they were specialized vocabulary. At other times we guessed that the word was archaic or belonged to a different dialect. We ended up with quite a few words without glosses. So you will need to make a decision whether to gloss the words during the workshop or afterwards. If the participants gloss the words they themselves have collected, you will get a gloss for every word. However it appears that you will collect more words if the participants spend their time thinking of words rather than defining them. There are four basic strategies for accomplishing the goals of this workshop:

1. Have the participants gloss the words when they have finished a section. This enables them to work almost entirely within the language and keeps them thinking in the language. Then when they have finished a section, they should go back through each domain, giving a gloss or short definition for each word.
2. Have the participants gloss the words after they have finished collecting the words in all the domains. You run the risk of taking the entire time collecting words and having to arrange to have the words glossed later. But you will be more likely to finish all the domains and collect more words.
3. Identify some participants who know the national language and have them gloss the words during the workshop. This way if they don’t know the word, they can ask the person who gave it. This is something you can do with people who aren’t good at collecting words. It gives them something important to do.
4. Plan to gloss the words after the workshop.

I used to recommend that the workshop participants gloss the words during the workshop. However now that about 200 languages have used the DDP word collection method, we have enough experience to say with some degree of confidence that it is best not to try to gloss the words during the workshop. People get far greater result by concentrating on collecting words. In fact several groups havecollected as many as 20,000 words in a two week workshop by concentrating on collecting words. If you do not speak the language, a gloss may be necessary for you to monitor their work and catch problems such as literal translations of English idioms. So you may need to have the participants gloss the words for the first day or two. If you do decide to gloss the words during the workshop, the glossing procedure is as follows: Translate the words into English (or whatever the national language is). Put the English right after the vernacular and separate them with a long dash. luganda — relationship bumanye — having met before mukago — friendship The English translation does not need to be perfect. It is just to give people a basic idea what the word means. After the workshop is over, the dictionary will be processed and expanded. Some of the steps involved in this process depend on the gloss. The gloss can be one word or a phrase. If the people don’t know what the equivalent English word is, just have them describe the meaning as best they can. You will work on writing good definitions later.

If there is time and energy, it helps to give the participants other tasks to do. I’ve found that the participants did not seem to get bored, perhaps because the domains kept them moving from one ‘topic’ to another. If they get tired of working on one thing, it may help to have them do something different from time to time. The following are possible suggestions. Many domains involve a scenario. The participants can write a brief description of the scenario. For instance the domain Sleep involves the following scenario: “When I get tired, I get ready for bed, go to bed, go to sleep, sleep all night, and wake up in the morning.” Even a short description like this contains many words and idioms. that belong to the domain. A detailed description will contain even more. Write expository texts about cultural topics that would use the words from a single domain. For instance someone could explain marriage customs, the naming ceremony, or traditional legal proceedings. For instance someone could write a description of a birthday party: “When we are going to have a birthday party, we send out invitations. Sometimes we will have a surprise party. The guests all bring a birthday card and present for the birthday boy or girl. During the party we play games. We have a birthday cake with candles. Someone lights the candles and then everyone sings Happy Birthday. The person whose birthday it is makes a wish and blows out the candles. Then he opens his presents.” This kind of description is a gold mine of cultural features, lexical items, collocations, and grammatical constructions. The other major technique for collecting words is to make a word list from a text corpus. If there is no tradition of writing in the culture, there may be no literature to form a text corpus. You can use the workshop to begin collecting texts. Texts will be very important later in investigating how words are used. I do not recommend that the participants write example sentences or polished definitions. It takes a great deal of training and practice to do these tasks well. You will simply waste your time. It is far better to base example sentences on real examples taken from texts. It is also better to write definitions based on textual evidence and careful introspection. So spend your time writing out descriptions of scenarios and cultural practices. Have the participants think about how they would organize the list according to their worldview. What would they choose as major categories? Which domains were hard to understand or didn’t fit their cultural domains? How would they reorder or reorganize the list to make it more logical? This would help me in improving future editions of the list to make it more universal. Future editions of these materials will include instructions for investigating the meaning and usage of the words in each domain, and expanding the word list into a full dictionary. If you hold a workshop, please send a report to me at the following address. I would be very grateful for any corrections or suggestions for improving the materials. Ron Moe 13440 N 44^th St Apt 2090 Phoenix AZ 85032-6365 U.S.A.