Experiments in language

Northcote Thomas experiments in language

Linguistic research formed an important part of Northcote Thomas‘ anthropological surveys in Southern Nigeria and Sierra Leone. Prior to the early 20th century, most research into West African languages had been undertaken by Christian missionaries. In the context of the emerging colonial sciences, an understanding of local languages was not only useful in terms of communication with local populations, but it also served the project of mapping ‘tribal’ or ‘ethnic’ groups, their territories and their historical relation with one another.

The languages people speak and the tribal or ethnic group names they are given were often used interchangeably. In this respect, Thomas introduced a more nuanced distinction between language and ethnicity. The titles of his published reports therefore refer to the ‘Edo-speaking’ and ‘Igbo-speaking’ people of Southern Nigeria, rather than, for instance, ‘the Edo’ or ‘the Igbo’. Alas, this recognition that language and ethnicity are quite different entities was not reflected in the subtitle of his Sierra Leone report: ‘The Timne and Other Tribes’.

Northcote Thomas, Specimens of Languages from Southern Nigeria
Northcote Thomas’ own annotated copy of his Specimens of Languages from Southern Nigeria (1914). Cambridge University Library, Special Collections.

Collecting specimens of language

Methodologically, Thomas’s anthropological surveys in West Africa between 1909 and 1915 were defined by practices of collecting and documentation. Thus, he collected ‘specimens’ of language in much the same way as he collected ‘specimens’ of material culture or, indeed, specimens of local botanical species. The use of the term ‘specimen’ carries an implicit assertion about the ‘scientific’ status of the anthropological surveys and the knowledge they produced, with its connotations of objectivity, rigour and authority. (Qualities that can, of course, all be contested.)

Northcote Thomas linguistic tour, Edo dialects, ,1909-10
Pages from one of Northcote Thomas’s linguistic notebooks, comparing dialectical differences in Edo-speaking areas of Nigeria. SOAS Library, Special Collections. (Click image to enlarge.)

The process of collecting linguistic specimens included the compilation of word lists, phrases and stories. For this, Thomas enlisted the assistance of interpreters. Finding reliable interpreters was a considerable challenge and there is much correspondence on this issue in the Colonial Office archives, especially relating to Thomas’s initial tour as Government Anthropologist in 1909-10. We learn, for example, that Thomas regarded the first interpreter who had been assigned to him – a schoolteacher named Erumese – as ‘reckless and inaccurate’, while he was frustrated that his replacement – a Corporal Nimahan of the Police Force, who was ‘thoroughly competent’ – was obliged to return to his police duties after a period of four months.

Excerpt from Anthropological Report on the Edo-speaking Peoples of Nigeria, Part II: Linguistics, in which Thomas lists the names of the interpreters employed during his 1909-10 tour, including Erumese, Corporal Nimahan, Osidora, Ogbebo, James Smart, George, Oganna and Isuma.

Thomas named these interpreters and acknowledged the extent and importance of their contributions in his Edo report. Unfortunately, in his subsequent reports, individual assistants are not named, though there is no doubt that their contributions remained vital. The role of interpreters also went beyond providing linguistic assistance. In a letter sent in 1911, during his second tour in what was then Awka District, for instance, Thomas praised his ‘junior interpreter’, one Alfred Nwile, remarking that he has displayed ‘great intelligence and skill’ in collecting botanical specimens.

The actual ‘collecting’ of words, phrases and stories, whether by Thomas or his assistants, was done either through direct transcription into text or with the use of a wax cylinder phonograph recorder. In appendices to his Edo Report, Thomas provided guidance notes for colonial officials, including use of the phonograph in linguistic documentation, and advice regarding language transcription. He provided a list of 150 words and phrases for translation to allow for comparison across languages, as well as more detailed questions about language usage. These were effectively the same techniques that Thomas and his assistants used during the four anthropological surveys.

Northcote Thomas, linguistic notebook, Igbo dialects, 1910-13
Pages from one of Northcote Thomas’s linguistic notebooks, comparing differences in Igbo dialects. SOAS Library, Special Collections. (Click image to enlarge.)

Recordings and transcriptions

Thomas wrote up and published the results of the linguistic research from the surveys in various books and articles. These included volumes of his main anthropological reports dedicated to ‘linguistics’, ‘vocabularies’, ‘grammar’, ‘tones’ and ‘dictionaries’, as well as separate volumes entitled Specimens of Language from Southern Nigeria (1914) and Specimens of Language from Sierra Leone (1916), which comprise of pages of tables of words translated into different local languages and dialects. These works were distributed to members of the colonial service, as well as to university libraries. How many people actually read them at the time is unknown – one suspects not many!

Northcote Thomas, Anthropological Report on the Ibo-speaking Peoples, Proverbs, Narratives, Vocabularies and Grammar
Pages from Thomas’s Anthropological Report on the Ibo-speaking Peoples of Nigeria, Part III (1913), illustrating Thomas’s method of phonetic transcription and comparing dialect differences between Onitsha, Awka and Bende Igbo. (Click image to enlarge.)

Duplicates of the wax cylinder sound recordings were also made available at the Horniman Museum in South London and the Pitt Rivers Museum in Oxford for scholarly consultation. Again, these seem to have been little used. The recordings have now been digitized by the British Library and we have been working with these throughout the [Re:]Entanglements project. In particular, we have been taking the recordings back to the communities in which they were recorded over 110 years ago, and it has been wonderful to witness as people listen to the voices of their ancestors and reconnect with this aural heritage.

In many cases, Thomas published transcriptions of the audio recordings, and it is fascinating to reunite these sounds and texts.

Experimenting with tones

Edo, Igbo and Temne are all tonal languages, in which lexical or grammatical meaning is altered by the pitch contour in which words are spoken. Thomas’s anthropological surveys took place at a time when the science of phonetics was becoming established in universities in Europe. Thomas was a friend of the phonetician Daniel Jones, who ran the Experimental Phonetics Laboratory at University College London. Jones had developed a method for determining what he termed phonetic ‘intonation curves‘ using phonograph cylinder recordings. Jones and Thomas worked together applying this technique to document the tonal changes in the specimens of Igbo speech that Thomas and his assistants had recorded during his 1910-11 and 1912-13 tours. According to Jones’ biographers, Beverly Collins and Inger Mees (1999), this was a pioneering piece of research on tone languages.

Daniel Jones, experimental phonetics, UCL, 1918
Daniel Jones demonstrating the use of the kymograph, an instrument for recording air pressure variations during speech. Experimental Phonetics Laboratory, University College London.

Thomas wrote up the experiment in Part VI of his Anthropological Report on the Ibo-speaking Peoples of Nigeria, providing transcriptions with musical annotations for some of the recordings they worked with. The specimens of Igbo language they worked with include such memorable expressions as ‘Does the goat frighten the dancer?’, ‘He took an egg, cried for a cloth, passed the bridge’ and ‘He put his foot on her waist and caused a big palaver’!

Northcote Thomas, Tones in Ibo, 1914
Top: Pages from Thomas’s Anthropological Report on the Ibo-speaking Peoples of Nigeria, Part VI (1914), providing a tonal transcription of phrases recorded in Awka Igbo dialect. (Click image to enlarge.) Bottom: The original wax cylinder recording from which the transcription was made, NWT 505 ‘Spoken sentences in Igbo’, recorded 1911 (British Library C51/2785). Thomas’s voice can be heard between the Igbo phrases giving the translation in English.
Top: Pages an article entitled ‘Some Notes on the Tones of the Ibo Language of Nigeria‘ (originally published in 1914), providing a tonal transcription of vowel sounds recorded in the Asaba Igbo dialect. (Click image to enlarge.) Bottom: The original wax cylinder recording from which the transcription was made, NWT 627a ‘Asaba vowel sounds’, recorded 1913 (British Library C51/2975). After Thomas’ introductory ‘ident’, the voice pronouncing the words is probably that of one of Thomas’ assistants, who also provides the English translations.

Orthographic debates

In his guidance for colonial officers, Thomas wrote that ‘For the collection of Vocabularies or native texts, two things are essential, one is, a certain amount of training of the ear, the other is an adequate system of transcription’. With regard to this system of transcription, he added, ‘the cardinal principles are, that each sound should have a sign peculiar to itself and that each sign should represent one and only one sound’.

At the time of Thomas’s surveys, there were a number of competing phonetic alphabets in use. Thomas used a system based on modifications to Latin script through diacritical marks. This was based on a Standard Alphabet devised by Karl Richard Lepsius for ‘reducing unwritten languages and foreign graphic systems to a uniform orthography in European letters’, published in the 1860s and recommended for adoption by the Church Missionary Society.

Northcote Thomas linguistic orthography and diacritical marks
Excerpts from Appendix A of Thomas’s Anthropological Report on the Edo-speaking Peoples of Nigeria, Part I (1910), setting out the orthographic system that he uses for different speech sounds. Thomas explained in some detail how it should be used and modified. (Click image to enlarge.)

In a review of Thomas’s Anthropological Report on Sierra Leone published in the Times Literary Supplement published in 1916, the reviewer criticized Thomas’s use of ‘inverted vowels and coined accents’, which he found confusing and wondered if there were not a more simple system. This provoked a lengthy exchange in the letters pages of the TLS that lasted seven months, in which numerous authorities debated various issues concerning phonetic spelling.

In Nigeria, the Lepsius system was superseded, first, by the adoption of a ‘Practical Orthography of African Languages’, developed by the International Institute of African Languages and Cultures in the 1920s, and, subsequently – in the case of Igbo – by the Ọnwụ system in the 1960s. The Ọnwụ system consists of 28 consonants and 8 vowel sounds.

The Ọnwụ system of orthography widely used in Nigeria today, with equivalent sounds as represented in the International Phonetic Alphabet in square brackets.

N. W. Thomas, linguist?

As may be discerned in the discussion above, Thomas was as much a linguist as he was an anthropologist. In 1914, while he was working in Sierra Leone, he was asked to advise on the introduction of linguistics in the training of new Colonial Service staff. Candidates who passed the examination were entitled to salary supplement. In the National Archives in Sierra Leone we discovered a draft paper Thomas had prepared entitled ‘Elementary Sketch of Phonetics’, which was evidently intended as a introductory text for teaching purposes.

Northcote Thomas Elementary Sketch of Phonetics
Handwritten manuscript of Thomas’s ‘Elementary Sketch of Phonetics’ in a correspondence file concerned with language training for officers in the Colonial Service, probably drafted in 1914. The manuscript includes annotated excerpts from proofs of Daniel Jones’ An Outline of English Phonetics (1918). Sierra Leone National Archives. (Click image to enlarge.)

In the event, it appears that this text was not adopted, and George Noel-Armfield’s book, General Phonetics for Missionaries and Students of Languages (1915) was used alongside reprints of the linguistic appendix to Thomas’s earlier Edo report. The latter was used as a guide for candidates who were expected to collect specimens of language from the colonial territories in which they served.

Thomas’s career as a government anthropologist came to an abrupt end in 1915 at end of his Sierra Leonean tour. He did, however, continue to write articles on linguistic themes, including a broader survey of what were then called ‘Sudanic languages‘ (languages of the Sahel belt) published in the Bulletin of the newly established School of Oriental Studies in 1920, and an attempt at reconstructing historical population movements through linguistic analysis in a paper entitled ‘Who were the Manes?‘ published the same year in the Journal of the Royal African Society.

Thomas also taught African languages, as an occasional lecturer at the Imperial Institute in London’s South Kensington, as part of the Tropical African Services Course. Candidates were evidently required to collect and transcribe language samples, as evidenced in a letter we discovered from Llewellyn Travers Chubb, sent to Thomas in 1925 from Bende in present-day Abia State.

Travers Chubb letter to Northcote Thomas regarding Tropical African Service course, 1925
Letter from Llewellyn Travers Chubb to Northcote Thomas, 21 February 1925, from Bende, Owerri Province, referring to Thomas’s teaching on the Tropical African Services Course in the autumn of 1924, and enclosing his word list assignment. (Click image to enlarge.)

Nothing of significance?

What are we to make of all this endeavour today? More recent linguists have been quick to dismiss the value of Thomas’s work. Betram Okolo, a linguist based at the University of Benin, Nigeria, argues that ‘nothing of significance’ was written on Igbo linguistics between 1890 and 1930, and describes Thomas’ efforts as ‘grossly inadequate’ and ‘misleading’. However, his remark that Thomas’ work ‘represents one of the most idle performances offered to the public on the Igbo language’ seems somewhat unfair. Indeed, it seems Okolo was not aware that the records on which Thomas conducted his tonal experiments were also recorded by him and his assistants over six years of fieldwork using primitive equipment in challenging conditions, or just how pioneering were his attempts with Daniel Jones at documenting tonal languages using ‘scientific’ methods.

Excerpt from Betram Okolo article, The History of NIgerian Linguistics
Excerpt from Betram Okolo’s article ‘The History of Nigerian Linguistics: A Preliminary Survey‘, published in 1981, in which he dismisses Thomas’s linguistic research, perhaps without fully appreciating its vast scope or pioneering nature.

While we might contest the assertion that Thomas’s linguistic work was an ‘idle performance’, its entanglement in the colonial project cannot, of course, be denied. Joseph Errington argues that ‘Colonial linguistics needs to be framed … as a nexus of technology (literacy), reason, and faith and as a project of multiple conversion: of pagan to Christian, of speech to writing, and of the alien to the comprehensible’ (Errington 2001: 21).

Furthermore, as Judith Irvine has recently noted, ‘These early projects contributed to the shape of African linguistics as we inherit it today, and – as part of the colonial enterprise – they had effects on the lives of the African languages’ speakers’ (Irvine 2008: 324). This is perhaps most evident in the use of (modified) European scripts to render many of Nigeria’s and Sierra Leone’s languages, and in the use of English as their national languages, such that younger people especially are turning away from their local languages.

Revisiting Thomas’s linguistic research

As part of the [Re:]Entanglements project, we have been collaborating with colleagues in the Department of Linguistics and Nigerian Languages at the University of Nigeria, Nsukka. In a future article, linguists Gloria Tochukwu Okeke and Ogechukwu Miracle Uzoagba will report on their experimental research on dialect change, comparing Northcote Thomas’s historical sound recordings with recreations of the same texts by present-day speakers of the same dialect. Their fascinating work suggests that the value of Thomas’s recordings may lie in the future rather than in the past.

Dr Gloria Okeke of the Department of Linguistics and Nigerian Languages, University of Nigeria, introduces her work exploring sound changes in the Awka Igbo dialect using Northcote Thomas’s historical sound recordings.

Selected references

  • Collins, B. and I. M. Mees (1999) The Real Professor Higgins: The Life and Career of Daniel Jones. Berlin & New York.
  • Errington, J. (2001) ‘Colonial Linguistics’, Annual Review of Anthropology 30: 19-39.
  • Irvine, J. T. (2008) ‘Subjected Words: African Linguistics and the Colonial Encounter’, Language & Communication 28: 323-343.
  • Okolo, B. A. (1981) ‘The History of Nigerian Linguistics: A Preliminary Survey’, Kansas Working Papers in Linguistics 6: 99-125.

Peripheral presences: N. W. Thomas’s Field Assistants

Assistant with umbrella. Peripheral presences in N. W. Thomas's anthropological photography. NWT 194b, 195a, 196a and 200a.
One of N. W. Thomas’s assistants, probably John Osakbo, as a peripheral presence at the edge of the frame in a series of physical type portraits photographed in Benin City in 1909. NWT 194b, 195a, 196a, 200a.

The image of the anthropologist as a heroic, lone fieldworker, battling through adversity in order to single-handedly document disappearing customs and rituals is a tenacious myth. Some anthropologists intentionally portrayed themselves in such terms. Malinowski‘s 1922 monograph, Argonauts of the Western Pacific, famously begins with the lines: ‘Imagine yourself, suddenly set down surrounded by all your gear, alone on a tropical beach close to a native village, while the launch or dinghy which brought you sails away out of sight’. In fact, we know that anthropological fieldwork is – and always has been – a highly collaborative endeavour. The important role of fieldwork collaborators – including fixers, brokers, assistants, interpreters and other participants – has, however, often gone unacknowledged. A notable exception was Franz Boas, who acknowledged his debt to his Tlingit-speaking assistant, George Hunt, who collected much of the data on which Boas’s publications were based.

N. W. Thomas was undoubtedly an energetic fieldworker, travelling extensively in Southern Nigeria and Sierra Leone in the course of some 55-months of anthropological survey work between 1909 and 1915. While Thomas was the sole author of the various reports and publications that resulted from this research, and is credited with assembling the thousands of photographs and sound recordings, and extensive collections of artefacts, botanical specimens and linguistic materials that are the legacy of these surveys, it is clear that this could not be the work of just one man. But who accompanied Thomas on his travels? How many assistants did he have? What roles did they play? One has to look hard to find a trace of such collaborators in the archives of Thomas’s surveys – but they can occasionally be glimpsed as peripheral presences.

This peripheral presence is most literally manifest when Thomas’s assistants appear at the edge of the photographic frame, holding a number board, supporting the photographic background sheet, or diffusing the sunlight with an umbrella. Many of Thomas’s photographic negatives are loosely framed, allowing peripheral detail to creep into the picture. The intention would have been to crop these images prior to publication, removing the traces of their co-production. As an experiment, such photographs can be differently cropped, placing the peripheral presences in the centre of the frame.

Peripheral presences. Details of N. W. Thomas photographs NWT 427 and 959.
Placing the peripheral presence in the centre of the frame. Again, probably John Osakbo, described by N. W. Thomas as ‘the most capable boy I ever saw’. NWT 427, 959.

Represencing Thomas’s fieldwork collaborators also entails recognising their trace elsewhere in the archive. In negative number NWT 261, a photograph of a group of Hausa musicians and dancers taken in Benin City in 1909, an assistant can be seen on the verandah making notes in what appears to be Thomas’s photographic register. These register books survive in the archives of the Royal Anthropological Institute and, indeed, the handwriting on these pages is not Thomas’s. Has the act of writing this very register entry been captured at the periphery of the frame?

Peripheral presences in N. W. Thomas's anthropological photography. NWT 261.
Left: A group of Hausa musicians and dancers, photographed in Benin City in 1909. Right: Detail showing figure on the verandah at the right edge of the frame, and of the photographic register entry he appears to be writing. NWT 261.

Within the photographic archive of the anthropological surveys, there are just five photographs of N. W. Thomas himself. These were likely taken by Thomas’s field assistants. In one intriguing pair of photographs, taken at the same location, it appears that Thomas and one of his assistants – probably Corporal Nimahan (see below) – have taken it in turns to photograph one another. This raises the question as to how many other photographs in the archive might have been taken by Thomas’s assistants rather than by Thomas himself.

N. W. Thomas (left) and an unnamed field assistant, possibly Corporal Nimahan, 1909.
Left: N. W. Thomas, possibly photographed by Thomas’s interpreter, Corporal Nimahan. Right: Unnamed assistant wearing corporal’s stripes (possibly Corporal Nimahan), probably photographed by N. W. Thomas. RAI 400.38267, 400.38292.

There appears to be only one entry in Thomas’s photographic register books in which it is noted that an assistant has taken a photograph. Thus photograph NWT 283 is described as ‘Burial of Legema, 26.3.09’. Evidently a sequence of four photographs was taken under this same number: 3 and 4 ‘by N.W.T.’, 5 and 6 ‘by John’. In fact we know a little more about ‘John’ compared with Thomas’s other assistants. This was evidently John Osakbo of Benin City. In a surviving letter from Thomas to the Colonial Office, sent from London in May 1910 after the completion of his first anthropological tour, Thomas requests that this assistant be paid a ‘retaining fee’ of £1 a month until his return to West Africa. Thomas describes John Osakbo as ‘the most capable boy I ever saw’, but notes that he was illiterate, and that the retaining fee should be paid on condition that he learn to read and write, and that he should also ‘receive training in photography’. It appears that Thomas’s request was granted. Thomas also recorded a phonograph of John Osakbo playing a song on a high-pitched woodwind instrument. Thomas’ voice can be heard at the start of the wax cylinder recording (NWT 16; BL C51/2164), ‘…song played by my servant, John, February 10th, 1909’.

 

It is likely that the number of individuals who accompanied N. W. Thomas on his travels varied from tour to tour. He travelled with camp equipment as well as photographic kit, phonograph and much else and would therefore have needed carriers. He seems to have travelled on foot, on bicycle and by hammock. In a letter to the Royal Botanic Gardens, Kew, written in 1928, recalling the collecting of vernacular names of plant species in Sierra Leone, Thomas mentions that Temne and Mende plant names were obtained from his hammock boys, and that they had been recruited in Freetown. Thomas relied on the assistance of interpreters, not only in his day-to-day interactions with people in the communities he visited, but also in compiling vocabularies and other linguistic data. In the preface to Part II of Anthropological Report on the Edo-speaking Peoples of Nigeria, concerned with linguistics, Thomas provides a list of the interpreters with whom he worked during his first tour and explaining the methodology he employed. Their names are: Erumese (Edo/Benin City), Nimahan (Kukuruku and Ishan), Osidora (Agbede and Kukuruku), Ogbedo (Edo/Benin City), James Smart (Sobo), George, Oganna and Isuma (Kukuruku). Nimahan was a corporal of the Southern Nigeria Police, and appears to have acted as both official interpreter and as representative of colonial authority. In Part III of Anthropological Report on Sierra Leone: Timne Grammar and Stories, Thomas notes that the first twelve stories published in the book ‘were recorded from the mouth of various members of my staff’.

In conclusion, by attending to their peripheral presences in the archive, it is clear that N. W. Thomas was not a lone fieldwork, but was accompanied and assisted in his anthropological survey work by an entourage of collaborators. While further work needs to be done to identify both the names and full range of activities they undertook, it is evident that their roles were fluid (‘hammock boys’, for example, provided ethnographic and linguistic information and did not simply transport the anthropologist on his itinerations). These collaborators were not peripheral to the anthropological project, but were in fact central to the endeavour. Hopefully, through the [Re:]Entanglements project, we will be able to identify more of N. W. Thomas’s Nigerian and Sierra Leonean collaborators, and correct the erroneous impression that Thomas was single-handedly responsible for assembling this remarkable ethnographic archive.