SUBSCRIBE NOW!! This first issue of CAANews is free. We intend to publish three more issues in January '85, April '85, and July '85. To receive them please send a check or money order for FIVE DOLLARS to Lee Sailer whose address appears at the end of this newsletter.
BONUS OFFER If there is sufficient material submitted (by you, of course), then we will produce more issues. So, in a sense, you can increase the value of your five dollars by encouraging people to contribute.
CALL FOR ARTICLES, NEWS ITEMS, ETC. It's very simple. Unless you submit material, we cannot print it. We hope CAANews will provide a forum where anthropologists discuss the relationship between anthropology and computing. A professor of mine used to say that, for interesting discussions to develop, first people have to be willing to disagree about some topic, and then someone has to be willing to say some thing worth disagreeing about. Then issue after issue, we can see how the ideas develop. Be assured that the editors do not intend to write a newsletter ourselves. In fact, we are willing to do the work because we expect to learn from the contributors. At present, we will consider any type of material on any topic, including product reviews, technical advice, tutorials, philosophical statements, and descriptions of software (commercially written or written by you -- available now or available in the future). At this point, even a letter describing the kind of material that you would like to see in CAANews would be useful. If you wish, call Dow or Sailer to discuss an idea for a submission, or send us something that you've already written, and let us summarize it. You will find our addresses later in this issue. Like I said, unless you submit it, we cannot print it.
If you haven't seen it, be sure to read Volume 6, No. 2 of Practicing Anthropology. It was almost all devoted to a large article called "Computer-Assisted Anthropology" prepared by Lee Sailer with the assistance of others. Most of the contributors were the people who attended the original meeting of the Committee for Computer Assisted Anthropology. The article covers 16 cases of computer-assisted anthropology ranging from field-note management to the simulation of market interactions. There is also a technical advice article written by Jim Boster that tells you how to pick your microcomputer for anthropological tasks. Practicing Anthropology is a publication of the Society for Applied Anthropology, 1001 Connecticut Avenue, NW, Suite 800, Washington, D.C. 20036, and costs only $10.00 per year.
It is amazing how many different ideas anthropologists have about what it means to "manage fieldnotes". I have talked to several people who are working on what one would expect to be similar projects, Gerald Gold at Toronto, Russ Bernard at Florida, Mike Agar at Maryland, and perhaps a dozen others. What stands out in my mind is that the systems have very little in common. Maybe this means that there are many different types of information being collected -- each requiring its own approach. Maybe there are just different types of anthropologist.
Whatever the reason, I think that all of us involved in the development of 'Fieldnote Management Technology' can benefit from open discussion of the issues. The computer-expert anthropologists need to discuss data-structures, algorithms, user interfaces, and other hardware and software topics. The computer-naive anthropologists (that is, those who will actually use FMT!) can contribute -- must contribute -- their ideas about what FMT should be able to do, and what it should be like to use it.
Since 1980, I have designed three FMTs, called Ethnos, UTS, and BBU. Because I had spent a year at Northwestern where I often talked to Ossy Werner about his work, Ethnos was designed to satisfy the needs of projects similar in nature to his Ethno- Medical Encyclopedia Project (Sailer, et.{al.{1984). Ethnos (Ethnographic Note System) currently exists in the form of a 2000 line Pascal program. I don't think it is usable without considerable revision, since its user interface is very primitive.
Ethnos was designed to help break a corpus of natural text, elicited from several informants, into very small, coherent fragments which could be reorganized by topics selected by the anthropologist. For example, as the reader of this newsletter, you might mentally extract out notes of some particular interest to you. By design, Ethnos is very sophisticated in the ways that it tries to help the user.
The lesson I should have learned from Ethnos is this: It is very hard to write computer programs that are very powerful, do everything that you can imagine, are easy to use, and WORK! I now often say, "It is easier to make a working program powerful than it is to make a powerful program work." To some extent, I bit off more than I could chew, given courses to teach, other research, family responsibilities, etc.
A different obstacle to the completion of Ethnos was (and still is) uncertainty. Look at it this way. Why should I invest large amounts of time on a project that may never be used? At the time that I was working on Ethnos, not a single member of my department, not even on paper, was "managing" fieldnotes at all, so far as I knew. Another contributor to my uncertainty was that I was unsure that Ethnos would be what people needed. I had already experienced the confusion of discussing my approach with other computer-expert anthropologists and having them say, "Huh??"
Then, in 1983 I worked with U. Pittsburgh psychologist, Charles Lidz, to design a system we called the Universal Text System. Why did we start from scratch to design a completely new system? Lidz, in contrast to Werner, is totally hostile to the notion of breaking natural text (transcripts of group therapy sessions) into pieces. To Lidz, each 2000 page transcript is a single, conceptually unified whole. Therefore, Ethnos was useless to him.
UTS allows the user to peruse a large corpus of natural text, marking 'points' of interest. Each point can be associated with any of a number of topics, or a other points in the text. Also, points can be associated with points in other texts, such as other transcripts, analyses, or even journal articles. UTS was designed to make it easy to find the points related to a certain topic, jump forward or backward a page or two, switch among several documents, find mutual occurrences of several related topics, and so on.
UTS may someday exist. Lidz has recently obtained funding for a project large enough to include software development. To date, no code has been written. UTS is designed for a multi-user computer in the $10,000 class, though it could possibly be used on a single-user micro with a hard disk (a Kaypro 10 or Morrow MD-11) costing about $3000.
I think UTS is a kind of fantasy system. Perhaps we think too small, but I don't see anthropologists taking $10,000 computers into the jungle. What we see are Osbornes, Kaypros, and some IBM PCs. Another thing I don't see is great commercial potential for FMT. If millions of people managed notes, we would long ago have seen software to help do it. I could be wrong -- perhaps people are just waiting for the right tools, but if I am right, then FMT will have to come from academia.
All this brings us to the third FMT I have designed, Basic But Usable (BBU). The most important difference between BBU and its predecessors is that it works! It is ready for people to take out and try. One reason for this is that it is very simple, though it accomplishes a useful task that you would be unlikely to attempt without it.
BBU is what computer scientists call a 'filter'. This means that it is a program that accepts input from a file, and converts that input into another, more useful form. BBU reads an input file of 'notes' and appends each note to one or more topic file. It also includes a title, date, and a list of other topics where the note is stored.
Functionally, BBU takes each note, which may be a few lines or several pages, and makes as many copies as necessary to file one copy under each requested topic.
To accomplish these goals, I applied the principle that a program should do one useful thing well. Rather than initiate yet another giant programming job, I picked a single task that would be immediately useful (I hope).
The next step is for anthropologists to use BBU, at least in an experimental mode. I do not consider BBU a commercial product. It is an intellectual experiment, and I hope that it will be treated accordingly. As with any new anthropological idea, BBU needs to be thought about, mulled over, and tinkered with. In time, BBU or its descendants might grow into another SPSS or VisiCalc.
To use BBU, use any editor that can create an ASCII file (WordStar, for example) to type in the notes. Each note is given a title, date and time, and a list of topics. An example --
/title Jose Ek Tzum biography
/date 26 June 1972
/topics pottery family markets Jose makes the best pottery,
according to Maxima. But he is not good at marketing it. I
think he takes the proceeds to buy rum. Nonetheless, his
family seems to prefer that he do the marketing...
/title ...
/date ...
/topics -family
etc. etc. etc.
After creating the file, the user (of a CP/M system) types
A>BBU file_name
Where 'file_name' is the name of the file created in the first step. Lines that start with '/' are control lines. The first note will be appended to files named pottery.ns, family.ns, and market.ns. The second note, along with the title and date, will be filed under pottery and market, but not family. Lines that start with '/' are control lines.
BBU is written in a standard dialect of C. It should run without change on any machine with a C compiler, including any CP/M machine, Apples, IBM PCs, most Radio Shack, PDP-11, VAX, etc. etc. etc. It currently runs properly using Manx Aztec C vs. 1.06 (not 1.05c!) and Mark Williams C. I will gladly share the source code with interested parties.
For those without C compilers, I can supply BBU.COM on almost any disk format, such as Osborne, VT180, and Rainbow. This program should run on most CP/M.
This section of the CAANews will be devoted to reports of what anthropologists are doing with computers. Please send the information on what you are doing, your successes and failures, and what you are interested in to Jim Dow. We will publish your reports with your name and address so our other readers can get in touch with you if there is a possibility of mutual assistance. The reports are going into a data-base that can be searched in the future if someone needs to find a person to help with a CAA problem. We will also publish comments here.
Please let us know if you would also like to participate in on-line conference to exchange this information. For this you need a terminal, a local network port, and at least $6.00 per hour.
I am particularly interested in the application of microcomputers to archaeological work. The development of reasonably reliable statistical packages (i.e. Daisy 2.0. Byte November 1983:560-570) makes it possible to actually use the micro in a field situation to assist in planning and controlling excavations that are on-going. At present, though, I must confess that the major use of the micro is for word processing. We have not yet found data base management to be as useful to us in practical terms as I had anticipated. We use VisiCalc and its clones in keeping track of numerical data which can then be analyzed with other programs. Our biggest problem at present is the difficulty of transferring data files from our Apple-II to the IBM mainframe here, although we have some promise of a solution as we devote more energy to the solution. So far, we have been pretty busy learning to use the micro to worry very much about the micro-mainframe connection problems. We have also had mixed results in using an Apple Graphics tablet as a digitizer. There seem to be some built-in errors and we have not had the time to work out whether these are due to hardware or software problems. I would welcome discussion on this issue with anyone who has used the Graphics tablet successfully for locational data recording.
To summarize: (1) Wordprocessing -- Many programs for micros are cumbersome and most are intolerably slow. I am using a package which is not so powerful but which has the advantage of speed in text entry [Letter Perfect]. For final printout of text, I am using another program which is slow as a snail, but very powerful [Screenwriter].
(2) Data entry and manipulation -- We find VisiCalc and its clones to be best for basic numerical data entry. We have been using Daisy 1.2.2 for preliminary statistical analysis. We have discarded other statistical programs because of major bugs, and plan to acquire the much more powerful Daisy 2.0 for future work. We have found less complicated DBM programs such as VisiDex to be more useful than more complicated programs, but plan eventually to go over to dBase-II.
(3) Digitizing -- We have used the Apple Graphics Tablet for digitizing data, but have found that measurements are subject to rather large errors in readings according to orientation, etc. At present, the tool does not seem to be precise enough for our purposes, although it is often handy for quick-and-dirty work.
Our work with the mainframe has been largely with SPSS and SAS. Many people in our unit are still using the mainframe and Script for wordprocessing. Contact person: Jon Muller, Department of Anthropology, Southern Illinois University, Carbondale, Illinois 62901.
Editor's Comment (JD)
Hardware does matter. I have never had any speed problems with word processing using WordStar on an Osborne, Kaypro-II, or IBM-PC. The IBM-PC and clones are already running SPSS micro-version. Perhaps the Apple, being a rather slow 6502-based machine, is not the best choice for serious research. I have been seeing good results with Kermit for micro-main file transfer. Kermit is a public domain program set that works on many main-frames and micros. It is released by the Columbia University Center for Computing Activities, Columbia Univ., N.Y., N.Y. 10027. Reader comment on the Apples as research tools will be welcome.
Morgan Tamplin at Trent University is presently using an Osborne-1 for word processing, statistical analysis, and programming special applications. He also uses a Commodore, Apple, Sinclair, and a VAX 11-780. He is keeping a list of Canadian archaeologists using microcomputers. Contact person: Morgan Tamplin, Department of Anthropology, Trent University, P.O. Box 4800, Peterborough, Ontario K9J 7B8, Canada.
Melvin Neville has both a Ph.D. in anthropology and a M.S. in computer science. His current research is in the area of primate behavior. He is interested in computer applications in this area and in studies of human evolution, biological bases of human behavior, human biological variation, animal behavior, primate anatomy, and evolutionary theory. In computer science his interests gravitate toward artificial intelligence and simulation methods. Contact person: Melvin Neville, Department of Sociology and Anthropology, Loyola University, 6525 N. Sheridan Rd., Chicago, Illinois 60626.
Program CONGRUENCE is a Fortran program for three-dimensional comparisons. Obtaining homologous planes of reference presents problems in comparisons of cartesian or polar coordinate measurements in three dimensions among specimens such as skulls or os coxae. Program CONGRUENCE calculates a transformation matrix for orienting one set of landmarks to those of another specimen. This matrix of cosines measures angular divergence among the right planes defining the coordinate space of the specimens. Cosines of angles among homologous pairs of landmarks are also produced after the pairs have been rotated to best fit.
A target may be used towards which other specimens are compared, or all specimens may be compared with each other. The procedure uses only those landmarks present for a given pair of specimens in any comparison. Printout includes new coordinates and measures of both average divergence of landmarks and average plane divergence among specimens. A listing of the program is available. The program was originally adapted for another use by David M. Heisler. Contact person: Robert Benfer, 210 Switzler Hall, Dept. of Anthropology, Univ. of Missouri-Columbia, Columbia, Missouri 65201.
The Department of Anthropology at Calgary has made a commitment to using computers in its anthropology program. Prof. W. C. McCormack needs programs for computer assisted instruction in anthropology, ethnography, cultural analysis, and semiotic analysis. Contact person: William McCormack, Dept. of Anthropology, University of Calgary, 2500 University Drive NW, Calgary, Alberta T2N 1N4, Canada.
I'm a student of Ossy Werner and am preparing my dissertation research. I own and Osborne-1 (double-density) and an Epson FX-80 printer, and last summer I took them to the field with me to experiment with computer-assisted ethnographic research on an archaeological field school in the Southwest. I studied staff interaction. Since my return I have been interested in two areas: (1) An extensive review of the literature on computer applications in the social sciences and humanities with an eye toward text analysis and DBMS (data base management systems) rather than statistics or highly structured data bases; and (2) the development of software or the adaptation of existing software for ethnographic DBM and analysis. Currently I am developing a DBMS and analysis system for my dissertation research. The hardware includes the possibility of a 15mb (15 megabyte) hard disk drive. I am attempting to hack out a program for reducing decision logic tables to flowcharts. My previous experience was primarily with a DEC PDP-11/70 on which I learned BASIC, FORTRAN, COBOL, PASCAL, and 8080/ASM. Despite Ossy's enthusiasm, I have not learned APL.
I am particularly interested in getting hold of names and addresses of other computer-oriented anthropologists and would like to receive papers on computers applications in ethnographic research which have not yet been published, such as those presented during the symposia held these last few years. I would be grateful for any contacts.
Northwestern U., where I am located, has one of the most computerized anthropology departments around. I am truly delighted to see that anthropologists are beginning to coalesce as far as computers are concerned. By the way, my ethnographic research interests have been primarily linked to computers. In 1981-1982 I did interviewing and observation with an academic data processing office and my proposed dissertation research will be among engineers in Silicon Valley. Contact person: Jeffrey Ward, Dept. of Anthropology, 2006 Sheridan Road, Northwestern University, Evanston, Illinois 60201.
Editor's attempt at an answer (JD)
As usual you seem to get what you pay for; however from my rather unsystematic perusal of the glossy pages I should say that Sanyo seems to be giving an awful lot of processing for the dollar these days. Morrow also has a cheap word processing package.My background includes 5 years experience as a consultant with SAS and IBM main-frame computers. I like the use of CMS/DMS for the creation of user-friendly data input fields on an IBM main-frame. Currently I am in charge of creating an in-house statistics/data processing system for the Hispanic Health Council in Hartford, Conn. The system is being designed for use by untrained community/agency personnel. The idea is to take quantitative analysis out of the university setting and make it available to a community agency. We have at present an IBM XT with STATPRO, Perfect Series Software (Writer, Calc, Filer, and Speller); and Peachtree General ledger. The system provides menu driven instructions from the moment it is turned on. It keeps the software on a hard disk and makes DOS as transparent as possible. The system is aimed at self-teaching the user how to use it. Contact person: Jeffrey Backstrand, Hispanic Health Council, 96- 98 Cedar Street, Hartford, Conn.
I have built and operated Heath/Zenith-89 and -100 microcomputers for out department to teach computer applications such as word processing and data base management. We use mainly CP/M-based programs such as Spellbinder and dBASE-II. I have written a text/graphics program in MBASIC called "Careers in Anthropology" that can be used to explain anthropology at highschool career days. Contact person: Alan Jacobs, Dept. of Anthropology, Western Michigan University, Kalamazoo, Michigan 49008.
I have four programs in BASIC for an IBM-PC personal computer. They are: (1) Measures of Polygyny, (2) Divorce Rates, (3) Population Change, and (4) Population Pyramid. The first two calculate very simple ratios. The third will figure growth rates and predict population, and the fourth draws am age-sex pyramid on a dot-matrix printer from data entered at the terminal. Contact person: James Vaughan, Dept. of Anthropology, Indiana University, Bloomington, Indiana 47405.
Frank Cancian has worked out a system for using SUPERFILE on an IBM PC to catalog, and retrieve field notes. Keywords are added to the margins of regular paper field notes. The keywords and other locational and time data are put in a regular ascii computer file and loaded into the SUPERFILE keyword data base. The locational data is retrieved by keyword searches. One advantage of this system is that it uses easily obtained hardware and software. He writes:
Keywords are written in the margin of the paper field notes. The input can be prepared on any word processor. The search is made on the basis of the keywords I have 200 keywords in a full dictionary of keywords used on my notes, and change them at any time. An entry can be a label for the page or anything else, including the actual text of the field notes, that you want to put between *C and *K. You can put up to 250 keywords on each entry, and key words can be up to 64 characters long. There is no practical limit on the total number of key words. When searching you can use single key words or combine them with operators AND, OR, and NOT up to a limit per run of 32 key words, or under some conditions 64. All the key word stuff is free field. You just need the // markers and the *K and *E In sum, it is a very flexible system. You can add to the new keywords when you decide to reclassify, and at little cost, etc. etc. I've recently produced an index of single keyword searches for all of my notes (about 300 pages) and keywords (200). It runs about 30 pages, and I plan to take it to the field with the spare copy of my notes when I go in a couple of weeks. More adventuresome types might take the computer.
Contact person: Frank Cancian, Program in Social Relations, School of Social Sciences, Univ. of Calif. - Irvine, Irvine, Calif. 92717.
I have become interested in the application of microcomputers in anthropology, either in the laboratory, field, or for teaching. A little over two years ago I obtained an Apple- II+ (64K, with CP/M card and 80-column card) though the Curriculum in Ecology. Aside from the use of WordStar in the preparation of manuscripts, we have started working on some software which is appropriate for the type of research I am interested in -- ecology and adaptation. And, I submitted a proposal for ecological work in Peru that includes the use of a microcomputer for management and analysis of data in the field.
We have completed a program called PCM 1.1. We are about to finish a second, similarly designed program, for calculating diet- breadth from optimal foraging models. We hope to continue and expand the use of the micro in this type of work. Contact person: Bruce Winterhalder, Curriculum in Ecology, University of North Carolina, Chapel Hill, North Carolina 27514.
PCM 1.1 is a user-oriented program for storing and managing time-series data and for calculation Colwell's parameters of predictability, constancy, and contingency. PCM 1.1 is written in Apple Pascal. It was developed on the following hardware configurations: Apple-II+ 48K microcomputer; Microsoft SoftCard and 16K Ram Card; Videx VIDEOTERM 80-column card; Mountain Computer CPS Multifunction card; 2 Apple 5 1/4 disk drives; a Zenith (ZVM-121) CRT; and an Epson MX-100 printer. It should run without modifications on any Apple-II+ with two drives, the CPS- Epson combination, and the Apple Pascal Language System.
Colwell, R. K. 1974 Predictability, Constancy, and Contingency of Periodic Phenomena. Ecology 55:1148-1155.
Winterhalder, B. 1980 Environmental Analysis in Human Evolution and Adaptation Research. Human Ecology 8: 135-170.
The National Science Foundation has set aside 7 million dollars for the purchase of supercomputer time for scientists who can use it. There also may be money to sponsor workshops on possible applcations of supercomputer work to anthropological research. The particular computers being made available are the CRAY-1 at the University of Minnesota, Minneapolis, the CYBER 205 at Purdue, Lafayette, In., and the CRAY-1/S at Boeing, Seattle, Wa. NSF will pay for CPU time, all other computer and peripheral chages, user support, and communications access costs. Information on the progam and facilities can be obtained from Lawrence Lee, Program Director for Supercomputing Centers, NSF, Washington, D.C. 200550 (202) 375-9717. Anthropology applications should go through Daniel Gross, Anthropology Program, NSF, Washington (202) 357-7804.