Substitution and Ellipsis in the Translation of English-Swahili Healthcare Texts

Even though substitution and ellipsis contribute to text cohesiveness just like any other cohesive device, the two grammatical cohesive devices are largely understudied. Focusing on the interplay between translation and cohesion, this study delves into the translation of substitution and ellipsis in Swahili healthcare texts. This study, theoretically anchored on Descriptive Translation Studies, set out to unravel the use of substitution and ellipsis and establish if there is any variation in the use of the two grammatical cohesive devices in the translated Swahili healthcare texts. The data used in the study is extracted from Orang’i (2020) doctoral study. Substitution and ellipsis are the basis for the manual comparison of the coupled pairs from the sample texts. The study has established that ellipsis and substitution are used sparingly in the texts. Though the use of substitution is limited, it emerged that clausal and nominal substitutions were prevalent. Equally, the use of ellipsis is almost non-existent in the Swahili healthcare texts though present in the source texts. The foregoing points to an endeavour by translators to make explicit that which is implicit in the source text and this led to the conclusion that explicitation is a norm in the translation of Swahili healthcare texts. The two cohesive devices are largely about one’s choice and can, to some extent, be avoided or minimally used. Overall, it was implied that substitution and ellipsis are not preferred cohesive devices in Swahili healthcare texts.


INTRODUCTION
Translated texts convey information that is expected to be unambiguously processed by the target audience. For the foregoing to happen, texts have to be cohesive. Therefore, cohesion is a fundamental aspect in texts and Dontcheva-Navratilova et al. (2017, p. 11) posit that cohesion is a textual property which fosters coherence as cohesive devices guide the reader in text processing. Halliday and Hasan (1976) came up with cohesive devices that they divided into grammatical cohesion (conjunctions, references, substitution and ellipsis) and lexical cohesion (collocation and reiteration). While it is appreciated that all the listed cohesive devices contribute to textuality, this article focuses on the grammatical cohesive devices of substitution and ellipsis in the translation of Swahili healthcare texts. The choice to jointly focus on the two is informed by Halliday and Matthiessen (2014, p. 635) who observed that substitution and ellipsis are variants of the same type of cohesive relation.
Even though the two cohesive devices contribute to a text's unity and make it coherent for the target readers just like references, conjunctions, and lexical cohesion, they are comparatively understudied. In fact, Tajeddin  (2017) observed that substitution and ellipsis are relatively neglected areas of research. This study goes against the neglect and gives focus to the two cohesive devices that have a blurred line differentiating them. Halliday and Hasan (1976), whose seminal work on cohesion has informed several studies, alluded to the two cohesive devices being the same process, in that ellipsis can be interpreted as a form of substitution. Toolan (2013, pp. 26-27) argues that the two can actually be referred to as partial ellipsis, which is substitution in this article, and full ellipsis, which is ellipsis in this article. On the other hand, Baker (1992, p. 187) noted that the two cohesive devices are grammatical as opposed to being semantic relationships. Kahigi (1997) conducted a study on the structural and cohesion dimensions of Swahili texts whereby substitution and ellipsis were given focus among other cohesive devices. He concluded that cohesive analysis is fundamental in the description of texts. Focusing on ellipsis, Mohamed (1986) did a contrastive study of Swahili and English discourse. He found out, among other findings, that translators of English into Swahili tended to frequently imitate English structure unlike the translators of Swahili into English who avoided imitations. Also, El-Shiyab (1998) discussed the impact of ellipsis in Arabic translations. The findings led to a conclusion that ellipsis is not a peripheral phenomenon but rather an important cohesive device in Arabic since it is essential between sentences and contributes significantly to the creation of texture and structure within texts. Still on Arabic, Izwaini and Al-Omar (2019) investigated the translation of substitution and ellipsis in Arabic subtitling. They concluded that repetition was the most dominant cohesive tie that is used as a counterpart of substitution and ellipsis in the English originals with explicitation as the main strategy. Another study by Arhire (2017) focussed on the translatability of ellipsis, substitution and references when they are enriched with stylistic, societal, and rhetorical values. The study concluded that explicitation and simplification are the appropriate translation strategies whenever the cohesive devices function only as formal links. It further points out that the cohesive system does not follow unique rules across languages and therefore cannot be transferred as such. The study has shed light on the translation of cohesive devices and that inasmuch as translators may try in their translation process, there cannot be uniformity in the use of the devices in the source and target text.
With the centrality of substitution and ellipsis in ensuring the cohesion of texts having been highlighted, it emerges that the two devices have not been given due focus particularly in the translation of healthcare texts. Translators of healthcare texts have to contend with a myriad of constraints in rendering texts to the target readers. Such texts will be falling short of their intended purpose if they are not cohesive since lack of cohesion affects the coherence of texts. It should also be borne in mind that healthcare texts, unlike other genres, do not give much leeway to the translator to make major adjustments in the target text that are lacking in the source text. This study, therefore, in an attempt to fill the void of no study on translated healthcare texts based on the English-Swahili language pair endeavours to: (i) describe the use of substitution and ellipsis in the translation of Swahili healthcare texts; (ii) establish any variation in the use of substitution and ellipsis in translated Swahili healthcare texts. Halliday and Hasan (1976, p. 88) define substitution as the replacement of one item by another. They further posit that it is a relation on the lexicogrammatical level, the level of grammar and vocabulary, or linguistic 'form'. It is worth noting that the substitute item has the same structural function as that for which it substitutes. Halliday and Hasan (1976) came up with three types of substitution: nominal substitution whereby the substitute one/ones always function as the head of the nominal group and can substitute only for an item which is itself head of a nominal group; verbal substitution with do as the verbal substitute in English which operates as the head of a verbal group, in the place that is occupied by the lexical verb and its position is always final in the group; and clausal substitution whereby what is presupposed is not an element within the clause but an entire clause. It uses so and not as substitutes.

Substitution
On the other hand, Toolan (2013), even though he prefers to call it partial ellipsis and uses substitution as an alternative, defines it as one in which some 'abridged' or condensed structure is used to stand in for the full sequence. Besides, Crystal (1995, p. 228) introduces the concept of pro-form, a word which replaces or refers to a longer construction in a sentence, as one of the ways in which a sentence can be shortened, to avoid saying or writing the same thing twice. He equally points out that substitution can also be referred to as a replacement. Baker (1992, p. 186) defines substitution as replacing an item(s) by use of another item(s). The most common cohesive devices used for substitution include so, do(es), did, one(s) and the same. It is imperative to establish how the translation of substitution is carried out in healthcare texts since languages have different linguistic norms.

Ellipsis
Ellipsis contributes to a text's cohesiveness and informs one in the interpretation of the intended meaning. Halliday and Hasan (1976) aver that ellipsis occurs when something that is structurally necessary is left unsaid, that is, there is a sense of incompleteness associated with it. Equally, Crystal (1995, p. 228) noted that ellipsis occurs when part of a sentence is left out because it would otherwise repeat what is said elsewhere. He adds another aspect on ellipsis that notwithstanding its reference to what has been said previously, it sometimes anticipates what is about to be said. Baker (1992, 187) defines ellipsis as one that includes omitting an item and replacing that item by nothing. It is commonly used to avoid repetition. Moreover, Halliday and Matthiessen (2014, p. 635) aver that it is through ellipsis that anaphoric cohesion is achieved, that is, we presuppose something by means of what is left out. Besides, they introduce the notion of ellipsis assigning prominence and non-prominence: if they are non-prominent (continuous), they are ellipted; if they are prominent (contrastive), they are present. It follows that if elements are left out in a text then they are considered to be lacking prominence. Ellipsis is also referred to as substitution by zero.
Just like in substitution, Halliday and Hasan (1976) categorise ellipsis into nominal, verbal, and clausal types. First, a nominal ellipsis is whereby a nominal group that is elliptical presupposes a previous one that is not. It involves the upgrading of a word functioning as deictic, numerative, epithet, or classifier from the status of modifier to the status of head. Secondly, verbal ellipsis refers to ellipsis within the verbal group. An elliptical verbal group presupposes one or more words from a previous verbal group. For a verbal group to be elliptical, it should not represent all its systemic features. Verbal ellipsis is further divided into lexical ellipsis and operator ellipsis. Lexical ellipsis is when a lexical verb is missing from the verbal group and it is taken as ellipsis from 'the right' since it involves an omission of the last word, which is the lexical verb. On the other hand, operator ellipsis happens from 'the left' and involves the omission of operators, that is, the initial element in the verbal group. The last type of ellipsis is clausal. It presupposes any clause in a complex and automatically does so to all the clauses that are contingent on it, that is, all that come after it (paratactic) and all that is dependent on it (hypotactic).
To sum up, it emerges that ellipsis is the exclusion of what has been said before informed by the possibility that it can be retrieved from context. It follows that if one cannot retrieve the information from the surrounding or context then meaning is lost and that leads to miscommunication. Like any other cohesive device, ellipsis, which sends one back to recover missing words, adds to a text's semantic structure. Halliday and Hasan (1976, p. 4) define cohesion as a semantic one in that it not only refers to relations of meaning that exist within the text but also those that define it as a text. Still focusing on a text, Hoey (1991, p. 3) presents cohesion as the way certain words or grammatical features of a sentence can connect that sentence to its predecessors or successors in the text. Equally, Munday (2016, p. 151) observes that cohesion is produced by the grammatical and lexical links which help a text hold together. In addition, Baker (1992, p. 180) posits that cohesion is the network of lexical, grammatical, and other relations which provide links between various parts of a text. From the foregoing, cohesive devices greatly contribute to making of meaning in texts and therefore determine, to some extent, the communicability of texts.

Shifts in the Translation of Cohesive Devices
Since languages do not have the same properties, it is expected that shifts occur whenever translation is done. Blum-Kulka (1986, p. 299) asserts that the process of translation necessarily entails shifts both in textual and discoursal relationships. Blum-Kulka goes further to argue that on the level of cohesion, shifts in types of cohesive markers used in translation seem to affect translations in one or both of the following directions: shifts in the level of explicitness, that is, the general level of the target texts' textual explicitness is higher or lower than that the source text; or shifts in text meaning(s), that is, the explicit and implicit meaning potential of the source text changes through translations.
It is not farfetched to say that shifts are norms in the translation process. Accordingly, translators in the mediation process make some changes that are necessitated by the expectations of the target audience, the morphosyntax of the language pair in the translation, translation strategies employed, patronage among others. It is a fact that target texts differ from the source texts largely due to the differences that exist between language pairs in a translation process (Baker, 1995;Lapshinova-Koltunski, 2015). On the same token, it is argued by Baker (1993, p. 188) that the translator need only be aware that there are different devices in different languages for creating texture and that a text hangs together by virtue of the semantic and structural relationships that hold between its elements. Further, the genre being translated presents its textual norms that a translator cannot ignore and thereby prompting shifts in the target text. Overall, it is imperative to bear in mind that translation is not and indeed cannot be an idiosyncratic process.

METHODOLOGY
This article is methodologically situated within Descriptive Translation Studies (DTS) with the healthcare texts as the products that are investigated. DTS, as propagated by Toury (1995), is ideal for this study since it provides for practical analysis of texts and does not prescribe how translations ought to be done. It eschews labelling translations as either good or bad. Further, it is target-oriented and speculation-free which points to the fact that: firstly, one is not given room to be subjective in opinion, and secondly, the source text is not given undue prominence as it was with prescriptive theorists. It is ideal for this study because it views translation as a communicative activity. Healthcare texts are meant to communicate. This article makes use of tertium comparationis, borrowed from contrastive analysis and articulated in detail by Kruger and Wallmach (1997) in the description of translations, as a basis for comparing the English and Swahili healthcare texts. The tertium comparationis in this article is substitution and ellipsis as illustrated in the Figure 1 below.
It should be noted that comparison is only done on sections that provide optimum opportunity to unravel the use of substitution and ellipsis in the healthcare texts under study. Toury (1995, p. 80) pointed out that: (1) every comparison is partial only: it is not really performed on the objects as such, only certain aspects thereof; (2) a comparison is also indirect in its very essence; it can proceed only by means of some intermediary concepts, which should be relatable to the compared aspect(s) of both texts; (3) these intermediary concepts should also be relatable to the theory in whose terms the comparison would be performed. Further, Toury (2012, in Munday 2016) proposes a three-phase methodology for DTS that incorporates a description of the product. It entails: (1) Situating the text within the target culture system, looking at its significance or acceptability.
(2) Undertaking a textual analysis of the ST and the TT in order to identify relationships between corresponding segments in the two texts. Toury refers to these segments as coupled pairs. This leads to the identification of translation shifts, both 'obligatory' and 'non-obligatory'.
(3) Attempting generalizations about the patterns identified in the two texts, which helps to reconstruct the process of translation for this ST-TT pair. The data in this article is extracted from Orang'i (2020) doctoral study and was collected from Nairobi County, Kenya. This study entails six texts, both in English and Swahili, on healthcare that focus on: cancer, malaria, HIV, AIDS, smoking, pneumonia. Total population sampling was used in the collection of the texts from health centres in Nairobi County, site of the study.

FINDINGS AND DISCUSSION
The study aimed to investigate the use of substitution and ellipsis in the translation of Swahili healthcare texts. This was done with an endeavour to understand how translators deal with the two cohesive devices that not only enhance the unity of texts but also contribute to meaning in texts. This section presents the findings and discussion based on twelve coupled pairs from the texts under study.
1 st couped pair: Cancer text ST: Cancer is usually not inherited. It is not passed from parent to child the same way that height and HIV are.

TT: Zipo baadhi ya saratani zinazofuata vizazi katika familia, lakini saratani kwa kawaida haiwezi kuambukizwa kutoka kwa mzazi hadi kwa mtoto jinsi ilivyo kwa kimo cha mtu. Kadhalika, saratani haiwezi kutoka kwa mama ikaambukiza mtoto kama ilivyo kwa ukimwi (HIV).
Whereas the ST has clausal ellipsis of passed from parent to child, the translator has not used ellipsis in the TT but has instead added an extra sentence that explains how cancer cannot be passed from the mother to the child. This an illustration of a translator's effort to make explicit what is deemed as implicit and may hamper the delivery of the intended message.
2 nd coupled pair: Cancer text ST: Radiotherapy is also used to kill or slow the growth of cancer cells. It can be used alone or with surgery or chemo. Radiotherapy can be given in two ways -from a machine outside the body or from objects put inside the body. Some people get both.

TT: Tiba ya miale (radiotherapy) pia hutumika kuua au kupunguza ukuaji wa seli za saratani. Inaweza kutumiwa peke yake au pamoja na upasuaji ama kwa pamoja na kemo. Tiba ya miale (radiotherapy) inaweza kutolewa kwa njia mbili -kutoka kwenye mashine iliyo nje ya mwili wa mgonjwa au kutokana na vifaa vilivyotiwa ndani ya mwili wa mgonjwa. Baadhi ya watu hupokea yote mawili.
The ST has substituted get radiotherapy from a machine outside the body and from objects inside the body with both and the TT too has substituted tiba ya miale (radiotherapy) mashine ikiwa nje ya mwili na vifaa kutiwa ndani ya mwili (from a machine outside the body or from objects put inside the body) with yote mawili (the two of them). These refer to the use of the cardinal numeral both and yote mawili (the two of them) in substitution. The use of the cardinal numeral in both the source and target text complements Mohamed (1986) assertion that Swahili tends to imitate the English structure in translations.
3 rd coupled pair: Cancer text ST: Radiation therapy isn't painful, but some of the side effects can be. If you have any pain, talk to your doctor or nurse. TT: Tiba ya miale haina maumivu makali, lakini baadhi ya madhara andamizi yanaweza kuleta maumivu. Unapohisi uchungu wowote zungumza na daktari au muuguzi wako. The ST has ellipted painful whereas the TT has not and uses maumivu (pain) which in this instance is taken to be an equivalent of painful. The kind of ellipsis in the ST is operator ellipsis, one that happens from 'the left' and involves the omission of operators, that is, the initial element in the verbal group, which falls under verbal ellipsis. 4 th coupled pair: Cancer text ST: Palliative care focuses on relieving suffering caused by cancer and helps people live as comfortably as possible for as long as long as they can. It is not intended to cure illness. TT: Huduma ya kukabiliana na makali ya athari za saratani hulenga kupunguza maumivu yanayosababishwa na saratani. Mbinu hizo husaidia watu kuishi bila maumivu kwa muda mrefu. Lengo lake si kutibu ugonjwa.
The ST passes the message in two sentences whereas the TT does so in three sentences. The TT has added mbinu hizo (those procedures) to refer to palliative care that is ellipted, nominal ellipsis, in the ST. The use of mbinu hizo (those procedures) in the TT is a substitution for huduma ya kukabiliana na makali ya athari za saratani (care to mitigate the effect of cancer). This is yet another instance where a translator strives to be explicit. 5 th coupled text: AIDS texts ST: If a husband has (or might have AIDS), should a Christian leader encourage him and his wife to use condoms when they have sexual intercourse? We think so. The spouse needs to be protected from possible infection. TT: Kama bwana anao (au anaweza kuupata), je kiongozi mkristo amtie moyo yeye na bibi yake kutumia mipira wanapofanya mapenzi ya kimwili? Tunafikiri hivyo. Yule mwenzi anahitajika kulindwa dhidi ya ambukizo. The ST has used so to substitute a Christian leader encouraging husband and wife to use a condom when having sexual intercourse and the same is replicated by the translator in the TT whereby hivyo (so) substitutes kiongozi mkristo kumtia moyo bwana na bibi yake kutumia mipira wanapofanya mapenzi ya kimwili (a Christian leader encouraging husband and wife to use a condom when making love). In this case, we have clausal substitution in both the ST and TT through the use of so and hivyo which are equivalents.
6 th coupled pair: AIDS text ST: We feel we Christians are not sinning if we encourage these people to protect themselves and others through condoms. If that is all we do, however, we are failing to give the Christian message of hope and deliverance from sin.
TT: Tunaonelea sisi kama wakristo hatutendi dhambi tunapowahimiza watu kujilinda wao wenyewe na wengine kwa njia ya mipira. Kama hilo ndilo tunalofanya tu, basi tunashindwa kutoa ujumbe wa kikristo wa tumaini na ukombozi kutoka kwa dhambi. In the ST, the verbal substitute do in the second sentence substitutes encouraging these people to protect themselves and others through condoms and hilo ndilo tunalofanya (that is what we are doing) in the TT substitutes kuwahimiza watu kujilinda wao wenyewe na wengine kwa njia ya mipira (urging them to protect themselves and others through the use of condoms To sum up, there has been evidence of shifts in the use of substitution and ellipsis and this agrees with Blum-Kulka (1986) who notes that shifts occur in the types of cohesion markers used in the target texts and records instances where the translator expands the target text by inserting additional words. The insertion of additional words is exhibited in instances where the source text uses ellipsis but translators opt not to use ellipsis in the target texts. The foregoing points to the use of explicitation in the Swahili healthcare texts. This finding is compatible with other studies that have found there is a tendency to use explicitation in the translation of ellipsis (Arhire, 2017;Izwain & Al-Omar, 2019). The translators do not want to burden the readers with the task of supplying meaning to what is not said in the texts. It is also possible that the target audience may not be in a position to retrieve the left-out part.

CONCLUSION
This study set out to investigate the use of substitution and ellipsis, which fall under grammatical cohesive devices, in the translation of Swahili healthcare texts. It was guided by two objectives: to describe the use of substitution and ellipsis in the translation of Swahili healthcare texts; to establish any variation in the use of substitution and ellipsis in translated Swahili healthcare texts. First, it has emerged that both the ST and TT make use of substitution albeit sparingly. It should be noted, however, that instances of substitution are so limited in the healthcare texts both in the ST and TT. The trend that emerged in the Swahili healthcare texts is that translators opt for equivalent pro-forms as used in the ST. Though instances of substitution were minimal, clausal and nominal substitutions were prevalent. Thence, substitution is not a preferred cohesive device in healthcare texts both in English and Swahili.
The use of ellipsis in Swahili healthcare texts is so minimal, in fact almost none, as opposed to its usage in the ST. The translators, going by the coupled pairs analysed, prefer to include any part that is left out or ellipted in the ST. This manifestation of the rare usage of ellipsis in Swahili healthcare texts, notwithstanding their presence in the ST, means that translators make explicit in the TT what is implicitly stated in the ST. It is against this backdrop that this study concludes that explicitation is a norm in the translation of Swahili healthcare texts. Translators endeavour to supply that which is left unsaid in the source texts and therefore make ellipsis a non-preferred cohesive device in healthcare texts.
Generally, the use of both substitution and ellipsis in healthcare texts is not prevalent and therefore indicates that they are not preferred cohesive devices and this agrees with Ambi (n.d,) who observed that the reason for the least occurrence of substitution and ellipsis in texts is because they are features of spoken discourse. Equally, the crucial information contained in healthcare texts motivates translators to leave nothing unsaid and that partly explains the rare use of ellipsis. It is therefore upon translators to be cognisant that cohesive devices are arguably inevitable in any text and should strive to achieve cohesiveness without compromising the central message.

RECOMMENDATIONS
Based on the findings, the study makes the following recommendations: First, the study focussed on the translation of grammatical cohesive devices of substitution and ellipsis underpinned by Descriptive Translation Studies (DTS), the same can be applied to study the other cohesive devices as put forth by Halliday and Hasan (1976).
Secondly, this study made use of manual analysis and it will be insightful if a similar study of a larger sample is done using corpus-based analysis.
Lastly, a study on the coherence of the Swahili translated healthcare texts is recommended to have a glimpse of how the target audience perceives them.