Sunday, 14 November 2010


At my desk in Utengule: shoebox in the foreground
Last weekend my daughter asked me if I had any empty shoe boxes: she needed one to fill with odds and sods for an annual charity appeal. There is of course a pile of old shoe boxes in one corner of my study, waiting for the day when I abandon computers and manufactured filing systems in favour of the cheap and cheerful homemade solutions of yore. My most treasured shoe boxes lie at the bottom of the pile. One of them is filled with the slips of paper on which I recorded my Sangu (shisango) dictionary in the field; the other with subject and person indexes of my field notes compiled in the first month or so after my return to Cambridge in early 1982.

I indexed my chronologically written and organised notes at the suggestion of my research supervisor, Ray Abrahams, and having never word-processed them remain eternally grateful for this simple piece of advice. Compiling a dictionary using a card index or slips of paper was a time-tested practice, as afficianados of the making of the Oxford English Dictionary and James Murray's 'quotation slips' will know. Alison Redmayne, my ethnographic mentor, suggested I use index cards for recording Sangu when I first visited her in Oxford in April 1980. She also later recommended that I get hold of a copy of Wilfred Whiteley's 'Suggestions for recording a Bantu language in the field', an article based on a brief study of the Fipa language, in which the same method is suggested  ("It is probably easiest to begin by collecting a word-list on cards", 1964: 2). At the same time she sent me photocopies of the slips of paper on which she'd made her own small collection of Sangu vocabulary in the 1960s. And afterwards she wrote "If you do not have a filing box and cardboard index cards you can make your own index cards and use a biscuit tin - that is how my Kihehe dictionary started" (letter from Oxford dated 11 July 1980). While writing this note I asked Alison for more details: she recalls quartering quarto-sized sheets of paper to make her own cards or rather slips. For many years now she's kept these in a metal index card file drawer. This is the only copy of her Hehe dictionary, a language which very few Europeans can speak like her.  

I ended up with a shoe box. I've forgotten now whether I took this with me to Tanzania in July 1980 or obtained it when I was out there. New shoes were available in the country in 1980 but like a lot of basic goods were in short supply. Stationery, however, was essential to the functioning of the socialist state, and I had no trouble getting hold of pads of plain paper slips. Indeed there are still some unused pads in my Sangu shoe box with the price marked on them: four shillings each. Once I was installed in the village of Utengule-Usangu I set about recording Sangu vocabulary on these slips. At the same time I was learning Swahili, the primary language of my research in this increasingly polyethnic area. But collecting Sangu words and phrases provided me with endless pleasure. Even on the leanest of days I was sure to hear or elicit new items that I scribbled down on scraps of paper before transferring them to the thin slips of paper that were filed away in the box on my desk in alphabetical order within appropriate morphological categories. Let me hasten to add that I didn't study the language as systematically as I might if I knew what I do now, or my research depended on a professional description of it. To make matters worse, I've got cloth ears and struggle to hear vowel-length or distinguish tones and stress, and I didn't mark any of these. Indeed I didn't really understand the pitch accent system of Sangu until the mid-1990s (when I had time to work through and build on the work of my predecessor in Utengule, the White Father-turned-linguist Jacques Bilodeau), and a later attempt to elicit the finer points of tense and aspect in the language fizzled out in frustration. But my rough and ready dictionary was adequate for everyday ethnographic purposes, including the sociolinguistic analysis of Sangu greetings and titles that became one of the chapters of my thesis (Chapter 5, "The theory and practice of misreading greetings", 1984: 126-157).

After I'd written my thesis on the Phoenix mainframe computer in Cambridge, I typed up most of my Sangu dictionary (Walsh 1985), though I didn't get round to including all of the linguistic information from the shoe box and other notes that I'd made on language use, including transcriptions of songs and other tape recordings. And although I've since done further work on particular parts of the lexicon, including Sangu plant and animal names (e.g. Walsh 1995; 1996), I haven't made any attempt to incorporate additional terms and definitions in the dictionary, which I no longer have in electronic form (except as a scan of the computer printout). Nonetheless, I'm pleased that the contents of my old shoe box have been of some use to subsequent researchers, including linguists working for the Summer Institute of Linguistics (SIL) (e.g. Tlustos 2000). For some years SIL researchers have used their own electronic Shoebox, "a computer program that helps field linguists and anthropologists integrate various kinds of text data".  As the blurb says, this programme "is especially useful for helping researchers build a dictionary as they use it to analyze and interlinearize text. The name Shoebox recalls the use of shoe boxes to hold note cards on which definitions of words were written in the days before researchers could use computers in the field." Such is progress, but I'd hate to be parted from my own tatty box of linguistic history.


Bilodeau, Jacques 1979. Sept contes Sangu dans leur contexte culturel et linguistique. Elements de phonologie du Sangu, langue Bantou de Tanzania. Textes des contes avec traduction et notes. Thèse de Doctorat de Troisième Cycle, Université de la Sorbonne Nouvelle, Paris.

Tlustos, Martin 2000. Draft Sangu dictionary (incomplete). Unpublished ms., Mbeya, December 2000.

Walsh, Martin. 1984. The Misinterpretation of Chiefly Power in Usangu, South-west Tanzania. Unpublished Ph.D. dissertation, University of Cambridge.

Walsh, Martin 1985. Shisango Dictionary. Unpublished ms. (computer printout), Cambridge, June 1985.

Walsh, Martin 1995. Snakes on the Usangu Plains: an introduction to Sangu ethnoherpetology. East Africa Natural History Society Bulletin 25 (3): 38-43.

Walsh, Martin 1996. Fish and fishing in the rivers and wetlands of Usangu. East Africa Natural History Society Bulletin 26 (3/4): 42-47.

Whiteley, W. H. 1964. Suggestions for recording a Bantu language in the field. Tanzania Notes and Records 62: 1-19.


  1. Where can one obtain a copy of Whiteley's "Suggestions for recording a Bantu language in the field"?

  2. Paste this link into your browser to download it from the online database of Tanganyika Notes and Records (note that it is almost 17 MB in size):