Along the stereotyping road: nineteenth and early twentieth centuries narratives of ukuhlonipha

1 Introduction

One day […] the missionary, whose name was Green, happened to say that the Cape gooseberries which the natives brought for sale were too green to use. Instantly the native women corrected her, declaring that she should express herself in a roundabout way and say the gooseberries were not ripe, for Green was her husband’s name, and she should never mention it (Kidd 1904: 237).

Not mentioning one’s husband’s surname^[1] is one of the many rules included in the customs of ukuhlonipha^[2] – also referred to as isihlonipho – among Nguni-language and Sotho-language communities of South Africa, Eswatini and Lesotho. Ukuhlonipha are a set of avoidance-based customs of respect applied to in-laws and, to some extent, to chiefs or kings. Ukuhlonipha traditionally involves context-dependent strategies of linguistic avoidance (e.g., father-in-law or mother-in-law names), physical avoidance (e.g., spouses and their in-laws), spatial avoidance (e.g., cattle pen) and material avoidance (e.g., tools) for both women and men (Mncube 1949; Raum 1973; Soga 1931). Nowadays, the custom is not observed to the same extent but still has significance in modern societies (Irvine and Gunner 2018).

In this paper, I intend to illustrate the deep historical roots surrounding the conceptualisation of ukuhlonipha, as suggested by Irvine and Gal (2000), by questioning the 19th and early 20th centuries published accounts of the custom in colonial-era South Africa and imperial-era Europe, written in English. I argue that a fair amount of descriptions of ukuhlonipha have been produced by colonial agents in the South African context, yet only a handful of detailed accounts made it to nineteenth century philological and anthropological theories in Europe. The knowledge circulation of ukuhlonipha across continents is evidence of a dichotomy of ethnographic interests, between the necessity to describe readable colonial subjects in the South African context, and the emerging necessity to establish universal knowledge strongly tied to theories of language in the European context (Deumert 2020). Consequently, the custom of ukuhlonipha stands out as a complex web of sociolinguistic and sociopolitical implications reflecting the social and political stakes of colonisation in British-ruled South Africa as well as the intricacies of knowledge as power in far-off imperial-Europe.

Produced for the colonial administration, for the missionary audience or for the general public, I investigated early exogenous published accounts of the custom of ukuhlonipha written in the South African context from the 19th century to the early 20th century. Observing a tendency to describe a linguistic custom of the women, as contemporary research about the custom also tends to focus on, I engaged in the imperial route to question the ways in which knowledge about ukuhlonipha had been circulating from the South African colonial context to the academic theories of imperial Europe.

After providing a brief literature review of ukuhlonipha, I outline the dominant knowledge circulation about the custom in the 19th century. In the next section, I introduce a four-part thematic analysis of missionaries’ writings. Firstly, I focus on a sociolinguistic account of ukuhlonipha produced for the administration. Secondly, I explore the accounts of missionaries describing some difficulties they encountered and attributed to the custom of ukuhlonipha. Thirdly, I suggest how dictionaries can illustrate the evolution of conceptualisation of ukuhlonipha. Fourthly I introduce two accounts of ukuhlonipha that contributed to counter the tendency to gender-stereotyping the custom. The last section intends to demonstrate how the feminine conceptualisation of ukuhlonipha nonetheless circulated from the South African context to European academic circles.

2 Ukuhlonipha in the literature

Before the mid-20th century, the knowledge authorities over the custom of ukuhlonipha could be categorised as such: the “from-experience” colonial agents accounts (19th and early 20th centuries), the “for-generalisation purpose” accounts of the Western academics and intellectuals (19th century), and the early 20th century “ethnographic-oriented” accounts of anthropologists and linguists. Around the mid-20th century, a new interest in ukuhlonipha by South African scholars is displayed in Faye’s early study of ukuhlonipha and clicks in Zulu (1925), Mncube’s (1949) thesis about linguistic ukuhlonipha as applied to Zulu and Xhosa women, or Kunene’s exploration of ‘hlonepha’ among Southern Sotho (1958).

It is interesting to note that the 19th century accounts of ukuhlonipha tended to focus on the custom as applied to Xhosa communities while most of the early 20th century studies focus on ukuhlonipha among Zulu communities. However, in the 1980s, Xhosa communities were at the focus of ukuhlonipha research, with Finlayson’s study of ‘isihlonipho sabafazi’ (1978, 1982, 1984, 2002) and its sociolinguistic exploration by Dowling (1988), or the work of Herbert on the social significance of the custom (1990b) and the sociohistory of its linguistic components (1990a). Interestingly enough, a new turnover can be noticed at the beginning of the 21st century since most of the recent publications about ukuhlonipha account for the custom among Zulu communities (Irvine and Gal 2000; Irvine and Gunner 2018; Luthuli 2007; Neely 2021; Rudwick 2008, 2011, 2013; Rudwick et al. 2006; Rudwick and Shange 2009; Zungu 1997), alongside a handful of work about ‘hlonepha’ or ‘ho hlompho’ among Sotho communities (Fandrych 2012; Thetela 2002, 2006, 2013).

Most of the studies of ukuhlonipha produced in the late 20th century and early 21st century observed a custom decay and have foreseen a rapid fall of ukuhlonipha (Dowling 1988; Herbert 1990b; Irvine and Gal 2000). However, it is now clear that ukuhlonipha is still a component of the Nguni-speaking communities, both as a custom still observed in some contexts (Fandrych 2012; Irvine and Gunner 2018; Rudwick 2008, 2011; Rudwick and Shange 2009), either in its traditional understanding in rural context or in its modern understanding in urban settings, but also, as I want to suggest, as a custom slowly turning into a cultural heritage artefact. Therefore, historicising knowledge of ukuhlonipha can contribute to shed light on past understandings of the custom and therefore reflect on contemporary perceptions and questionings of ukuhlonipha, as it still affects contemporary lives, at least by the general scheme of social behaviour governed by respect strategies (Irvine and Gunner 2018).

3 From ‘British Kaffraria’ to Oxford: a hegemonic knowledge of ukuhlonipha

A thorough survey of ukuhlonipha accounts published in English in the 19th and early 20th centuries established a restricted knowledge circulation based upon the earliest explicit account of the custom. From 1850 onward, dozens of descriptions of ukuhlonipha were published in South Africa, but none had such an influence in European academic circles as John W. Appleyard’s in The Kafir Language (1850) had.

In 1850, the Wesleyan missionary John Wittle Appleyard^[3] was the first to explicitly^[4] describe the custom of ukuhlonipha in his grammar The Kafir Language (1850). Focusing on describing linguistic features of isiXhosa, Appleyard introduces the custom as one of the characteristics of the extended verba corpurum he observed in the language. In this manner, he introduces the custom as a women-only and linguistic-only “national custom” of the Xhosa. “It may be mentioned as a remarkable fact, that the Kafir women have many words peculiar to themselves. This arises from a national custom, called ukuhlonipa, which forbids their pronouncing any word, which may happen to contain a sound, similar to one in the names of their nearest male relatives. [ …]” (Appleyard 1850: 70).

Seemingly brief and not much detailed, this account of ukuhlonipha is nonetheless the first to state that the Xhosa people – or as suggested by Appleyard, the Xhosa women – have a linguistic pattern of avoidance involving names of relatives and being governed by a national custom. Interestingly enough, right before this description, Appleyard describes another linguistic pattern of avoidance that he apparently does not consider as part of the “national custom called ukuhonipa”.

There is a difference observable amongst some of the Kafir tribes, in reference to the use of certain words, arising from a kind of superstitious objection which they feel against employing a word, that is similar in sound to the name of one of their former chiefs. Thus, the Amambalu do not use ilanga, the general word for sun, because their first chief’s name was Ulanga, but employ isota instead. For a similar reason, the Amagqunukwebi substitute immela for isitshetshe, the general term for knife. (Appleyard 1850: 69–70).

However, one can be stricken by the similarity between the prohibition of phonetically similar words to the names of family members and the prohibition of phonetically similar words to the names of chiefs or former chiefs. In his grammar, Appleyard had in fact produced two different accounts of a similar custom, the custom of ukuhlonipha, applying both to family members and political figures. Nevertheless, what Appleyard labelled as “ukuhlonipa” was a linguistic behaviour of women, therefore rooting the first written knowledge of the custom in a gendered conceptualisation that would be further reused, as he was an influential figure in imperial 19th century.

From Appleyard (1850) short linguistic definition of ukuhlonipha to European academic circles, knowledge about the custom quickly travelled overseas and appeared from 1863 in the Lectures on the Science of Language of Prof. Friedrich Max Müller.^[5] In Müller’s introductory lecture, the custom of ukuhlonipha is introduced alongside the Tahitian custom of pi’i,^[6] to demonstrate how what could initially be labelled as an “accident” or a “fancy peculiarity” of a language could in fact emerge in geographically and linguistically distant languages (Müller 1863: 45). Moreover, while Appleyard did not link the “superstitious objection” to the custom of ukuhlonipha, it should be noted that Müller used both points to support his theory of women’s influence on language, stating that if men were bounded by the custom with regards to their chiefs, it was clear evidence of a “mere feminine peculiarity […] [that] extended its influence” (Müller 1863: 48).

Acclaimed professor of Comparative Philology and Comparative Mythology at Taylor Institution in Oxford, Müller’s account of ukuhlonipha, privileging a gendered conceptualisation of the custom, quickly became the authority over European academic writings using ukuhlonipha to support their anthropological, philological, historical or linguistic claims.

Overall, as soon as Appleyard’s account reached Europe, it became the primary source of most of the academic writings of the 19th century discussing the custom of ukuhlonipha, but only behind the authority of Müller’s Lectures on the Science of Language (1863). In fact, anthropologists or historians such as Edward B. Tylor (1865), John Fiske (1872) or Andrew Lang (1884) and philologists or linguists such as Archibald H. Sayce (1875) and later on Alice Werner (1905), attributed their knowledge of ukuhlonipha to Müller’s description and theorisation of the custom.

Nevertheless, the description of ukuhlonipha in Tylor’s Researches into the early history of mankind and the development of civilization (1865) suggests that Appleyard or Müller were not the only figures to contribute to create knowledge of the custom in the academic world. Indeed, Tylor’s account of ukuhlonipha (1865), while following on Müller’s, also introduces A Zulu-Kafir Dictionary (1857) from Jacob L. Döhne’s, missionary to the Berlin Mission and American Board Mission, as one of his sources. Therefore, were Appleyard’s and Müller’s depictions of ukuhlonipha the only accounts that circulated? While it surely dominated the construction of knowledge – at least in the academic context – a fair amount of accounts describing ukuhlonipha have been published in the 19th and early 20th centuries.

4 The coloniality of knowledge: ukuhlonipha narratives from South Africa (1811–1915)

Missionaries dominated the 19th century narratives of the custom of ukuhlonipha. Whether on linguistic work such as grammars (Appleyard 1850) and dictionaries (Colenso 1861; Davis 1872; Döhne 1857; Kropf 1899), on personal life accounts (Brigg 1888; Shaw 1860) or on monographs relating the history and folklore of the Southern African people (Fleming 1854; Holden 1866; Leslie 1875; Shooter 1857), or even on papers for and by the administration (Warner 1858), missionaries tremendously contributed to construct knowledge of ukuhlonipha. While each account could be taken separately, it is interesting to note how they – almost – all account for a similar narrative, one that encourages the substitution of the traditional society in favour of a ‘civilised’ Christian society, and therefore strongly wishes for the traditional customs to disappear (Mills 1995).

To explore the representation of ukuhlonipha under missionaries’ pens, I offer a thematic order rather than a chronological order. Given that knowledge of ukuhlonipha does not seem to have circulated much within the colonial context of British-ruled South Africa, it is relevant to reflect on similar missionary narratives that might have been published years from each other, but that nonetheless describe the custom for the same purpose. With this in mind, we will start with one of the earliest accounts of ukuhlonipha, written for the administration by the former Wesleyan missionary John C. Warner, and published in Colonel Maclean’s A Compendium of Kaffir Laws and Customs (1858). Secondly, we will focus on the abolitionist narratives of ukuhlonipha through the life accounts of the Wesleyan missionaries William Shaw (1860) and Arthur Brigg (1888), and in the monograph of William C. Holden (1866). Thirdly, we will focus on the accounts of ukuhlonipha in dictionaries published throughout the 19th and early 20th centuries, especially because they are evidence of the evolution of representation with regards to ukuhlonipha and the traditional customs in general. To bring the analysis to a close, we will introduce the early account of Joseph Shooter (1857), chaplain to the British settlers in Durban, along the writings of David Leslie (1875), a knowledgeable hunter close to the Zulu royal family, to illustrate the gender-stereotyping tendency surrounding knowledge of the custom of ukuhlonipha.

5 An imperial circulation of knowledge: ukuhlonipha between stereotypes

In the first part, I introduced the main circulation of knowledge identified in the 19th century accounts of ukuhlonipha, from Appleyard’s The Kafir Language (1850) to Müller’s Lectures on the Science of Languages (1863). In the second part, I have shown that a fair amount of accounts has been published throughout the 19th century, displaying descriptions of the custom as applied to women and as applied to men. However, in this part, I want to demonstrate how the circulation of knowledge, both within and beyond the South African context, contributed to embed a feminine conceptualisation of ukuhlonipha that spread until the 21st century (Irvine and Gal 2000; Irvine and Gunner 2018).

The ambiguity of knowledge surrounding the custom of ukuhlonipha is simultaneous to the first account of the custom being published. Indeed, as stated earlier, Appleyard’s description of ukuhlonipha is divided in two: an explicit depiction of “ukuhlonipa” as a linguistic custom of women, and a separate remark about a “kind of superstitious objection” supposedly applying to the whole community (Appleyard 1850: 69–70). Therefore, if one only takes description in which the custom is labelled as “ukuhlonipa”, it becomes an object of knowledge as a linguistic custom of the women. However, when Müller’s (1863) uses Appleyard’s description, he took both remarks into account and suggested that ukuhlonipha was a custom applying to women within the family, but also to men with regards to the chiefs (Müller 1863: 47). Thus, ukuhlonipha entered the European academic circles as a general custom of the Nguni-language people, further divided into a set of gendered rules. However, the initial confusion was not yet to be erased. Indeed, while Tylor’s account follows on Müller’s, he introduced the political ukuhlonipha towards chiefs’ names, but then stated that “it is also among the Kafirs that the peculiar custom of uku-hlonipa is found” (Tylor 1865: 147), dividing the political ukuhlonipha from what was then understood as the custom of ukuhlonipha, meaning a familial linguistic ukuhlonipha applying to women as Tylor proposes by quoting Döhne’s account (1857). In a similar vein, Fradenburgh’s account (1874) operates the same division between political ukuhlonipha and familial ukuhlonipha of women labelled as “ukuhlonipha” (Fradenburgh 1874: 39–40). Furthermore, other academic accounts of the custom inspired by Müller only introduce ukuhlonipha as a linguistic custom of the women (Fiske 1872: 223; Lang 1884: 84–85) to compare it with the Tahitian custom of pi’i, a custom that ironically enough resembles the rules of the political ukuhlonipha and not of the familial ukuhlonipha – as only noted by Lefèvre’s comparative account of ukuhlonipha and pi’i in the Encyclopédie Générale (1871: 247).

While Appleyard’s account cannot be easily defined as a gendered or ungendered description of ukuhlonipha – because they are both at the same time – academics nonetheless decided to underline its feminine aspect. But what can be said about the accounts that were explicitly non-gendered? Warner’s account in A Compendium of Kaffir Laws and Customs from Colonel Maclean (1968 [1858]) and Leslie’s account in his conference “The Native Custom of ‘Hlonipha’” published in the posthumous book Among the Zulus and Amatongas (1875), were both the primary sources of a handful of academic accounts about ukuhlonipha. Moreover, Shooter’s account of men and ukuhlonipha also became a primary source for the heavily referenced The Golden Bough of James Frazer (1900) and both the Natural history of man: Africa and The uncivilised races of men in all countries of the world of John G. Wood (1868, 1872), therefore slightly contributing to counter the gender-stereotype behind narratives of ukuhlonipha in the academic landscape.

In 1879, in the same year and by the same author, Warner’s in Maclean (1968 [1858]) and Leslie’s (1875) detailed descriptions of ukuhlonipha finally reached European intellectual circles. James Anson Farrer, jurist specialised in the philosophy of justice, published a monograph, Zululand and the Zulus (1879a), and an essay, Primitive manners and customs (1879b), in which each account of the custom of ukuhlonipha is based on a different primary source. Farrer’s contribution does add to the ambiguity of knowledge stated above, for he describes ukuhlonipha as a custom applying to women and to men in his monograph (1879a: 124–126), and as a linguistic custom of the women in his essay (1879b: 233–234).

In fact, the dissimilarity in Farrer’s accounts somewhat originates from his primary sources. In Farrer’s monograph, the primary source is the fine-grained account of David Leslie, who introduced a three-part ukuhlonipha, “the family, the tribal, and, in the case of the Zulus, the national” (Leslie 1875: 172), detailing how both women and men are expected to behave according to ukuhlonipha. Leslie’s account was also relayed in the same way by the Scottish jurist John F. McLennan in Studies in Ancient Histories. Second Series posthumously published in 1896 and in James Frazer’s Golden Bough (1900) – which had Fleming (1854), Shooter (1857), Maclean (1968 [1858]), Leslie (1875), Theal (1886) and Kranz (1880) as its primary sources. In Farrer’s essay, the primary source is both Leslie’s account and the early account written by John C. Warner. Farrer probably eclipsed the men part of ukuhlonipha to follow the path of most of the comparativist academics who focused on comparing pi’i and ukuhlonipha, since he introduces ukuhlonipha to support the similarity between ukuhlonipha and a linguistic custom of the “Caribs of the West Indian Archipelago” (Farrer 1879b: 233). Hence, Farrer appears to be the only author to have been influenced by Warner’s account without suggesting the ukuhlonipha rules of the son-in-law. In fact, both John Lubbock, a British archeologist turned ethnologist and later politician, and Frederick Starr, an American anthropologist and geologist, provided a non-gendered explanation of ukuhlonipha based on Warner’s account, in their respective essays The origin of civilisation and the primitive condition of man (1882: 14) and Some first step in human progress (1895: 235).

Overall, the custom of ukuhlonipha is either described as a women-only custom or as a custom having a great deal of implications for the women and few implications for the men. Even when the accounts offer a little explanation about ukuhlonipha as applied to men, it always appears more as a suggestion and less as a fact. Yet, ukuhlonipha was indeed a custom applying to men as it has recently been shown by Irvine and Gunner (2018) but it seems to have been erased from European academic theories of the 19th century.

To conclude, I want to underline the language stereotype that is concomitant to the gender stereotype. In fact, from all the above accounts and their interpretations, one can only notice how ukuhlonipha is depicted as a linguistic-only custom. However, the earliest description of ukuhlonipha identified in Alberti’s Description physique et historique des Cafres, sur la côte méridionale de l’Afrique (1811) only introduces some non-linguistic implications of the custom such as the interdiction placed upon women with regards to cattle (Alberti 1811: 129) or the avoidance behaviours expected from father-in-law and daughter-in-law and from mother-in-law and son-in-law (Alberti 1811: 135). Moreover, in the colonial context of South Africa, accounts of the custom as implying specific social behaviours are reckoned in almost half of the publications, varying from a single sentence to suggest a physical avoidance of in-laws, to a paragraph to describe some spatial avoidances or daily interactions sociolinguistically governed by ukuhlonipha. Yet, in the academic publications, only five accounts do offer both linguistic and non-linguistic implications of ukuhlonipha, relying on Shooter (1857), Maclean (1968 [1858]) or Leslie (1875) as primary sources. In fact, the linguistic stereotype seems to account for the dichotomy between ‘frontier linguistics’ and ‘imperial linguistics’, as suggested by Deumert (2020). Nineteenth century Western accounts were embedded in a universal knowledge dynamic and ukuhlonipha could therefore be used as an example of the linguistic ‘peculiarities’ found around the world, as to support some civilised versus uncivilised language theories. Nineteenth century accounts from South Africa were embedded in the colonial administration’s knowledge dynamic and illustrated the desire to fully understand people to turn them into “readable subjects” (Deumert 2020: 182), therefore, ukuhlonipha had to be understood as sociolinguistic rather than linguistic-only because it could not erase the social and political consequences of the custom within British-ruled South Africa.

6 Conclusions

The custom of ukuhlonipha entered the English-speaking knowledge dynamic as a linguistic object, in the writings of missionary linguists of the 1850s. Simultaneously, more personal writings such as monographs and life accounts offered another early narrative that included both the ‘hlonipha of speech’ and the ‘hlonipha of action’ – to use Bryant’s terminology (1905). Nonetheless, the knowledge circulation from South Africa to Western intellectual circles indicates that the linguistic aspect of ukuhlonipha was the centre of attention, especially with regards to women. Considering the amount of knowledge written about ukuhlonipha, both as a linguistic and as a sociolinguistic custom throughout the 19th and early 20th century, this study illustrates the strong historical roots behind the folk rationalisation of ukuhlonipha suggested by Irvine and Gal (2000: 47). Moreover, looking further into ukuhlonipha accounts offers an opportunity to (re)establish the significance of the custom during the colonial encounter, underlining how narratives of the custom can reveal a sense of cultural resistance over Christianity while also underlining how women could exercise agency by their observance of ukuhlonipha. Furthermore, given the linguistic-only and the women-only tendencies in the exogenous accounts of ukuhlonipha, it is necessary to engage in a discussion around ukuhlonipha with the underrepresented endogenous accounts produced by Nguni-speaking authors and informants (cf. James Stuart Archive; Mayaba 1972; Mncube 1949; Mzamane 1962; Soga 1917, 1931; Van Warmelo Collection).^[9] It would indeed contribute to reveal the concomitant creation of knowledge in colonial-era South Africa, and further on in 20th century South Africa, and would perhaps expand, nuance, refute or even support the dominant narratives surrounding the custom of ukuhlonipha.