Contrast, Contact, Convergence? Afrikaans and English Modal Auxiliaries in South African Parliamentary Discourse (1925–1985)

This article investigates modal auxiliaries in original and translated Afrikaans and South African English parliamentary discourse in the period 1925–1985. Against the background of the sociolinguistic history of language contact in the bilingual South African parliament (1910–1994), it analyses (a) the contrastive differences in the use of modal auxiliaries in South African English and Afrikaans, (b) potential cross-linguistic influence in the use of modals between the two languages, and (c) the way in which contrastive differences and cross-linguistic influence are reflected in translations. In both languages, modal auxiliaries are more common in parliamentary discourse than in general usage. There is little evidence of overall convergence; there are, however, cross-linguistic similarities in specific pragmatic uses of modals in parliament. Translations show a large degree of shining-through from the source text, alongside adjustment to target norms; the tension between these two forces is variable, and influenced by social factors.


Introduction
Contrastive pragmatics is concerned with the comparison of how different languages are used to perform similar functions in similar situations, often with teaching or translation applications in mind (Verschuren, 2016).The focus is on those communicative aspects that may give rise to misunderstanding due to variability in the use of linguistic resources (House, 2006: 249).In settings where different languages are in contact, these differences in the use of pragmatic resources may lead to cross-linguistic influence (Verschuren, 2016); findings from contact linguistics clearly demonstrate that such influence takes place at all levels of language, including (and even especially) the pragmatic (see e.g.Matras, 2009).Prolonged contact may result in changes in the use of one language under the influence of another language, a possibility that Kranich (2016) investigates for translation-induced pragmatic change in German, and for which she reports qualified support.In this context, translationinduced change is a (non-prototypical) form of (indirect) contact, where translators are the main agents of contact, and the contact occurs largely in the written mode, from there disseminating to non-translational usage (see e.g.Bisiada, 2016;Kranich, 2014Kranich, , 2016;;Malamatidou, 2016;Neumann, 2011;Pang andWang, 2020 -Kotze, 2021 provides a comprehensive overview of the state of the art).
In evaluating the methodological approaches in contrastive pragmatics, Verschuren (2016) points to the challenge of determining whether speech acts taking place in two different languages are sufficiently equivalent to be contrasted to establish the pragmatic differences between the languages.Kranich (2016: 18) therefore argues in some detail for the equivalence of the written documents in English and German that she compares.Likewise, Kádár and House (2020: 142-151) explain how the notion of a ritual frame provides a meaningful context of comparison through delineating standard situations within which data from different languages can be sampled and compared.
To resolve this methodological problem, the ideal situation is one where two different languages are used in the same communicative situation.Such data is difficult to come by,1 but will go a long way to resolving the methodological challenge of comparability.In this article, we analyse data from one such situation: the bilingual parliament in South Africa , which spans the period from unification until the first democratically elected representative government (see Section 3).In this time period, both English and Afrikaans2 were used by parliamentarians, without an interpreting service.Thus, during this period of time in the parliamentary history of South Africa, the conversation across the parliamentary floor took place with Afrikaans and English alternating as part of the same conversation.Certain speakers consistently used only one language, while some speakers used both languages, sometimes even in the same contribution from the floor (see Section 3).3 This dataset not only resolves the matter of comparability of speech acts to be analysed to contrast the pragmatics of two languages (in this case, Afrikaans and English), but also allows for the investigation of a unique contact setting, both synchronically and diachronically, thus creating scope for understanding pragmatic cross-linguistic influence between the two languages, and linguistic change, in real time.The nature of the contact is complex: the bilingual parliament of this time period reflects both institutional and individual bilingualism, and also incorporates bidirectional translation between the two languages.The South African Hansard until 1994 was published in two versions: a complete English and complete Afrikaans version, in which all presentations made in English on the floor were translated into Afrikaans for the Afrikaans Hansard, and conversely, presentations in Afrikaans were translated into English for the English Hansard.
1 Such situations seem rarer than intercultural contact situations between speakers of different languages taking place in a single common language (whether the native language of one of the conversation partners or a lingua franca that is native to neither).2 Prior to 1925, Dutch was used as language of record alongside English.Afrikaans only came into official use in 1925.3 Since the South African parliamentary discourse from 1994 and onwards changed dramatically in composition (English becoming the dominant language in parliament, and Afrikaans becoming one among several minority languages, only rarely used), the parallel analysis of Afrikaans and English in the enclosed space of the all-white parliament cannot be continued into the fully representative parliament of 1994 onwards.Thus, the possible future development of Afrikaans and English convergence in the kind of immediate contact situation, where all speakers were assumed to understand both languages, stops with the end of the 21st parliament, shortly before the general elections in April 1994.The language contact situation in parliament subsequent to this is material for a separate study; Kotze and Van Rooy (2020) present some initial steps in this respect.
The proceedings of these parliamentary discussions are published in edited form as the Hansard, the official record of parliament.The Hansard is thus not a verbatim record of parliamentary proceedings, but it nevertheless aims to give a reasonable representation of the proceedings, while serving as a fully adequate substantive record.There are differences between the official, written Hansard and the original spoken interactions, as shown by Hibbert (2003), Mollin (2007), and Kotze et al. (in press) for the South African, British and Australian Hansard respectively, particularly in that the Hansard tends to be more conservative, and not only rewrites the spoken language into an acceptable written form, but also does so in ways that are often aligned with older norms.While this is a potential limitation of the data, given that we are interested in the possible occurrence of pragmatically-driven language change, it also means that evidence of such change despite the conservative bent of the Hansard would be particularly convincing.
In our analysis, we focus on modal auxiliaries in the bilingual South African Hansard spanning the period 1925-1985.Our choice of modal auxiliaries as feature for the investigation is functionally motivated.Parliaments are sites of rhetorical contestation, but paradoxically also sites of communicative collaboration: in the legislative arm of government, politicians develop their positions in opposition to the alternatives proposed by other parties, but they also come together to ultimately decide on policies and accept legislation that direct the executive arm of government (Ilie, 2003: 73).Modal auxiliaries play an important role in parliamentary persuasion and argumentation (Simon-Vandenbergen, 1997;Vukovic, 2014), and as Kotze and Van Rooy (2020) show, the use of the English modals has indeed changed in frequency and function during the course of the 20th century in the English version of the South African Hansard (although they excluded the translated parts of the record and analysed only the speeches and debates originally presented in English).
As pointed out above, our corpus reflects the sociopolitical history of the period 1910-1994; the omission of data from the earliest and latest part of this timeframe is a practical matter, explained in more detail in Section 3 and Section 4. Our analysis focuses on (a) the contrastive differences in the use of modal auxiliaries in English and Afrikaans, (b) cross-linguistic influence in the use of modals between the two languages, potentially leading to convergent usage over time, and (c) the way in which contrastive differences and cross-linguistic influence are reflected in translations, in the two languages.
In Section 2, we focus in more detail on the choice of modal auxiliaries as feature for investigation, summarising existing research on changes in the use of modal auxiliaries in South African English and Afrikaans, and the way in which modal auxiliaries are approached in translation in the two languages, from a diachronic perspective.This sketches the necessary background against which our analysis of the use of modal auxiliaries in the parliamentary context, specifically, should be viewed.In Section 3, we provide more detail on this parliamentary context, sketching a brief outline of the history of parliament and Hansard in South Africa, in relation to the sociolinguistic history of Afrikaans and English in the country.Section 4 presents the methodological approach of the article, outlining the corpus composition, data extraction, and analytical steps.Section 5 presents and discusses the findings.We first provide an overall analysis of the frequency and use of modals in the two languages (original and translation) over time, considering both the frequency of modals overall (Section 5.1), and the patterns for semantic groups of modals (Section 5.2).Against the background of these general findings, we carry out a fine-grained analysis of usages that appear of particular pragmatic importance, focusing on modals expressing volition and prediction, in original speeches (Section 5.3).Section 6 concludes the article, by summarising the answers to the research questions and reflecting on the implications of this study for contrastive pragmatics more generally.

Modal Auxiliaries in South African English and Afrikaans
The change in modals in South African English,4 also reflected in the Hansard, should be understood against a broader canvas of change in the modals of South African English (Wasserman, 2014), and Afrikaans (Erasmus, 2019), and the possible role of language contact between English and Afrikaans in these changes (Wasserman and Van Rooy, 2014;Wasserman, 2016;Van Rooy, 2021).
The most important finding from Wasserman's (2014) diachronic study of modals in South African English  is that the changes have by and large been similar to those happening in other native varieties elsewhere in the world, except that the modals in general declined somewhat more in frequency and the quasi-modals in general increased somewhat less in frequency, leading to an overall decline in the frequency of the modal auxiliaries over time.There is one very clear exception in South African English: the modal must has not declined in frequency to the same degree that it has declined in other varieties of English.This is attributed to the influence of Afrikaans: the very high frequency of must in South African English relative to other varieties of English is matched by the high frequency of its cognate moet in Afrikaans, and there is also evidence of increased semantic similarity between the two modals over time (see also Wasserman, 2016).In addition, Wasserman (2014) identifies a sharp increase in the frequency of can from the middle to the end of the 20th century, not matched by other native varieties of English.For Afrikaans, Erasmus (2019) reports a strong increase in the frequency of kan, the cognate of can, continuing into the early 21st century -matching the finding of Wasserman (2014).Erasmus (2019) also reports a strong increase in the deontic uses of moet especially in the 1940s, whereafter the overall frequency and the strength of the obligative force of moet show a gradual decrease, to become less face-threatening, also matching the finding of the semantic change in must reported by Wasserman (2016).
This previous research on language change in the modals of Afrikaans and English is mainly in the form of comparisons between the available historical corpora -those of Wasserman (2014) for English and Kirsten (2019) for Afrikaans.In these monolingual corpora, the data were sampled from native speakers of the two languages in contexts of written language production addressed to other native speakers of the language (at least in the main).There is also one diachronic study on the translation of modals in South African English and Afrikaans, by Redelinghuys (2019).She compiled a bidirectional translation corpus of English texts with their Afrikaans translations and Afrikaans texts with their English translations, spanning the period 1910-2016.She identifies a number of specific cases where modals in Afrikaans and English show converging change over time in particular registers.However, she finds that translation follows the change in the original language production, and concludes that there is no evidence that translations lead contact-induced change; at most they reflect changing norms in the target language, but at times they act as a conservative force, and reflect and reinforce the norms of an earlier stage of the target language (variety).
Against this background, the three questions that this article sets out to answer are: 1.
What are the contrastive differences in how the modals of English and Afrikaans are used in the bilingual South African parliament of 1910-1994?2. Is there evidence of pragmatic change in the use of modals in these two languages in the context of the direct contact situation of the bilingual South African parliament, potentially reflecting convergence?3. How are these contrastive differences, and possible changes over time, reflected in translations in Afrikaans and in English?There are very clear pragmatic goals that modals help to achieve in parliament, including to persuade others of viewpoints, to assess the epistemic strength of evidence presented, and to impose obligations on the executive and the population at large.Within these pragmatic considerations, the changes in the use of modal resources can tell us whether the situation of direct contact sees more convergence in the use of modals than in language use in non-contact settings; in other words, whether the direct contact leads to a reduction of contrastive differences between the languages involved.

3
The South African Parliament (1910-1994): Contextualisation The South African context in general, and the South African parliament that we focus on here in particular, raises challenges to the 'natural' one-to-one mapping of language and culture, a 'linguaculture' that is often assumed in contrastive pragmatic studies.We therefore do not frame our analysis as one of, primarily, differences in 'cultures' , since the cultures involved are not sufficiently enclosed to be investigated as separate systems.The speakers in our dataset reflect variable degrees of bilingualism, biculturalism, or cultural syncretism; and speakers who might be associated with one 'culture' use both languages.In our dataset, thus, language and culture exist in fluid many-tomany mappings.
We treat this multiplicity as a given, and in the discussion that follows, we try to illustrate these complex intersections in the political history of parliament, contextualised within bigger dynamics of animosity and unification between linguistic groups, conflicts and alliances in party-political affiliations, and external socio-political and economic factors.We also reflect on the consequences that this complexity had on the record-keeping processes of the Hansard.

A Brief History of the Sociolinguistics of the South African Parliament
The bilingual contact situation surrounding the 20th century parliament goes back to the annexation of the Cape Colony by the English from the Dutch, temporarily in 1795 and permanently from 1806.5 Official Dutch monolingualism under Dutch colonial rule made way for an anglicisation policy under British rule from 1822 onwards, although it did not result in English monolingualism.Widespread societal bilingualism developed alongside a diverse set of official language policies in the different colonial polities of 19th century South Africa: English as official language in the British colonies of Natal and the Cape, with Dutch reintroduced as additional language of parliament of the Cape by 1882; and Dutch as official language in the two nominally self-governing Boer Republics of the Orange Free State and Transvaal.The relationship between the two linguistic communities of South Africans of European descent was strained in the 19th century, and culminated in the Anglo-Boer War of 1899-1902, in which the British Empire finally defeated the independent Boer Republics and merged them into the Union of South Africa with the existing British colonies.Yet, sufficient contact was maintained across the linguistic divide to ensure that a meaningful segment of both communities attained proficiency in the other language.
When the Union of South Africa was established, antagonism between the two camps was strong and the emotional consequences of the war still acute.One attempt to forge (white) nationhood was in the establishment of the South African Party, through the political cooperation of English and Afrikaner6 politicians, who jointly sought to develop a South African identity inclusive of both groups.This followed from an expanded view of Britishness on the side of English imperialists like Milner, to accommodate Afrikaners, and conversely, a willingness to seek common ground on the side of the Afrikaner leaders who were largely the same people that led the armies of the Boer Republics in the war, less than a decade earlier (Lambert, 2000).The initiative never managed to penetrate the entire Afrikaner community, so as early as 1914, the National Party was established as a break-away faction from the South African Party, with an exclusive white Afrikaner ethnic supporter base.Through coalition with the Labour Party, they gained power in 1924, but in the aftermath of the Great Depression of the early 1930s, a renewed attempt at consolidation was made with the merger of the South African Party and National Party into the United Party in 1934.A minority of National Party members did not join the United Party, but continued under the National Party banner as opposition to the United Party until they gained the electoral victory of 1948, redefining the United Party as mainly English party again, and consolidating Afrikaner support in the National Party (Giliomee and Mbenga, 2007;Maloka, 2001).
After the establishment of an independent Republic of South Africa, the National Party gradually enlarged its supporter base to all whites (Maloka, 2001), while the United Party declined with every election, ceding support to the National Party on the right, and to various break-away factions on the left -starting with the Progressive Party in 1959, which gradually grew bigger, through absorbing various other breakaway groups from the United Party, until it became the official opposition after the 1978 election, and the United Party itself disbanded in 1977 (Giliomee and Mbenga, 2007;Suzman, 1993).
While the United Party of the 1930s and 1940s, and later, and the Progressive Party, especially from the 1970s onwards, had Afrikaans public representatives, the majority of their parliamentarians were English speakers, and their Afrikaans representatives often used English in their speeches too, as seen in the Hansard data.By contrast, the National Party communicated almost exclusively in Afrikaans until the 1970s, when it started to elect more English-speaking representatives and when some of its Afrikaans-speaking representatives, particularly ministers, took to using English in some of their speeches and other responses during debates.While issues of language and culture were very prominent in the first half of the twentieth century -often under the banner of 'the two European races' , the black and white race issue became increasingly more important as the century progressed.There is a strong connection, but not absolute identity, between language and political position (left or right in the context of the time), such that parties with a majority of English representatives were to the left of parties that were exclusively or mainly Afrikaans.

Dutch, Afrikaans and English in South Africa
The two languages in interaction, Afrikaans and English, trace their origins to the European colonisers who transplanted them to South Africa.Dutch arrived earlier, in 1652, and changed more dramatically than English, speakers of which only arrived in substantial numbers in 1820.Until the 19th century, the older of the two colonial languages was called Kaaps Hollands, the Cape dialect of Dutch, and metropolitan Dutch was the language of literacy, education, the church and government administration.From the last quarter of the 19th century, there were calls for the recognition of the local vernacular as a different language, Afrikaans.These calls grew into a movement, which saw the first spelling rules and grammar of Afrikaans published in 1917, and official recognition in 1925 (Kirsten, 2019;Van Rooy and Van den Doel, 2011).
When the Union parliament was first established in 1910, Dutch was recognised as the second official language, alongside English, despite Afrikaans being the spoken vernacular of the non-English-speaking parliamentarians.Recordkeeping was done in Dutch until 1924, although the vernacular was increasingly reflected in the records of debates.Legislation was still drafted in Dutch, even in 1925, when the official languages act that recognised Afrikaans was van Rooy and Kotze Contrastive PragmaticS 3 (2022) 159-193 passed -the officially signed version of the act was written in Dutch.From 1926, all parliamentary records were in Afrikaans.A period of uncertainty about norms lingered well into the 1930s (Kirsten, 2019), and by 1925, the first sampling point for the present study, full standardisation had not been achieved in the formal written language.
'South African English' , as we use it in this article, refers specifically to the native variety of English spoken by descendants of the British Settlers, and individuals subsequently absorbed into the speech community (Bekker, 2012).It shows much more continuity with the metropolitan variety of English than Afrikaans does with Dutch, and strong British normative influence was still present in 1925, and lingered for most of the 20th century (Lanham and Macdonald, 1979).In the local context, this variety should more accurately and narrowly be described as White South African English, in contrast with other native varieties such as Indian South African English and nonnative varieties, including the second-language English of black speakers and Afrikaans speakers (Botha et al., 2021;Van Rooy, 2017).During the period under investigation in this study, there is little normative instability in South African English, with British norms remaining the official yardstick for spelling and grammar, unlike Afrikaans that had to establish its own independent norms after its 'declaration of independence' from Dutch.

3.3
The Language of the South African Hansard The history of the Hansard as record of parliamentary proceedings in South Africa is set out by Van Wyk (2010).The national Hansard officially only dates from 1910, with the Union of South Africa.Prior to this, the Cape Colony, Natal, and the two Boer republics all had forms of parliamentary record keeping; our focus here is only on the national Hansard produced from unification onwards.
No Hansard was produced in 1910, the year of the establishment of the Union of South Africa.In 1911, a committee investigated how a published Hansard should be produced.At this time, an official verbatim Hansard was deemed too expensive, and a decision was made in favour of a newspaper Hansard (i.e., a Hansard compiled from newspaper reports).Debates about the language of the Hansard were present from the beginning, with questions about whether there should be a single volume, with speeches and debates recorded simply in the language presented (a so-called 'piebald' Hansard), or a double volume, in English and Dutch (later Afrikaans).
These early Hansards, until 1915, are thus not verbatim (Van Wyk, 2010: 49), and are generally seen as of poor quality (Van Wyk, 2010: 45), because the Hansard tradition was not yet strongly established.From 1915From -1924 no Hansard was produced (it was deemed too expensive in the post-war economic conditions).In 1920, a parliamentary committee recommended the establishment of an internally produced Hansard, verbatim, in the language presented.This proposal was debated in subsequent years: language and cost remained key issues in deciding between a bivolume bilingual Hansard (relying on translation), and a monovolume bilingual Hansard (Van Wyk, 2010).
The 1924 Fourth Session of the Fourth Parliament is the only piebald Hansard.Subsequently, the decision was made for bilingual volumes in English and Afrikaans, with a substantially verbatim approach.This approach was continued for the entire duration of the period under investigation in this article.As is the case for reporting practices in other parts of the world (see Kruger et al., 2019), the practicalities of recording and producing the Hansard have changed radically over the years, moving from shorthand reporting by Hansard reporters, dictation to typists, and editing; to, later, the use of recordings and speech recognition.In the period under investigation in this study, reporting relied on the more traditional methods.
To our knowledge, there is little systematically documented information available about the translators of the Hansard in this time period, or their working methods (although Van Wyk, 2010 provides a wealth of anecdotes).However, in general, it may be anticipated that in this context, a tension between a strongly incentivised source-text orientation (remaining as close as possible to the original utterances in translation, given the function of the translation as an official record of parliamentary proceedings), and sensitivity towards developing target-language norms in the two language(s) (varieties) in question.Also impacting on the latter would be the variable relations of status between the two languages; it may be assumed that the ways in which status differences between languages condition source-text or target-language orientation (Toury, 2012) were also at play here (see Redelinghuys, 2019 for more detailed discussion).
In other words, translation can, in this context, be seen as a secondary language contact situation overlaid on the first language contact situation in the chamber itself -where the forces identified in the previous paragraph impinge on the ways in which that real-life, interactive contact situation is subsequently mediated for the written record.

Corpus
The corpus was built from the original Hansard records, obtained from a university library.We opted for a periodisation structure of intervals of decades, starting in 1925 (the first year in which fully parallel Hansards were created), and ending in 1985.By 1995, the composition of parliament had changed dramatically (see Footnote 3).The complete English and Afrikaans proceedings of every last day, or in case of a short sitting, last two days, of every calendar month of the selected years were scanned and converted to text format by means of Optical Character Recognition software.These were subsequently proofread by assistants,7 with samples re-checked by the authors, while adding mark-up to separate original and translated portions of the texts, and set aside lists, direct quotations and records of written questions and answers.
The size of the individual components of the original texts and translated versions (excluding all the lists, direct quotations and written questions and answers), is reported in Table 1.

4.2
Data Extraction and Analysis All modal verbs were extracted manually in the Concordancer function of WordSmith 8.0 (Scott, 2020).The extraction included contracted negated forms, as well as non-contracted negated forms that form orthographic units like cannot in English and moenie 'must not' in Afrikaans.All modals that required disambiguation were disambiguated manually by the authors (e.g. to separate the use of may as modal auxiliary from its use as proper noun to indicate the month of May).Where appropriate, normalisation of all frequency counts to a relative frequency per 100,000 words was done.
We opted to cluster the modals in the same way that Wasserman (2014) does, following the Longman Grammar of Spoken and Written English (Biber et al., 1999), into three semantic groups:

5
Results and Discussion

Overall Results
The results for all modals together, in the Afrikaans and English originals, and corresponding English and Afrikaans translations, are presented in Figure 1.The Afrikaans original texts consistently have more modal auxiliaries than the Afrikaans translations or English texts, original and translation, except for A χ² test of independence was used to evaluate the relations between the different datasets.For both these pairs of originals with their translations, the χ² test (with df=6 in all cases) fails to reject the hypothesis of dependent samples (χ²=7.64,p=0.27, for Afrikaans originals and their English translations; χ²=7.42,p=0.28, for English originals and their Afrikaans translations), while the χ² test rejects the dependence hypothesis of Afrikaans and English originals (χ²=45.35,p<0.001),Afrikaans originals and Afrikaans translations (χ²=69.09,p<0.001), and English originals and English translations (χ²=50.20,p<0.001).
The statistical analysis does not provide support for a somewhat idealised hypothesis that the use of modals in the original Afrikaans and English Hansard is similar over time, but does confirm that the shining-through effect of the source texts in translations is more pronounced than the degree to which translations approximate the quantitative norms of the target language.This is most clearly visible in 1935 and 1985, where the Afrikaans and English originals change in different directions compared to the previous sampling points.The English translations follow the rise of the Afrikaans originals in 1935, and the decline of the Afrikaans originals in 1985, and conversely, the Afrikaans translations follow, and is even more pronounced than, the decline in the English originals in 1935, and similarly follow the increase of the English originals in 1985.
The variable tension between fidelity to the source text and approximation of target-language norms alluded to in Section 3.3 is, however, also in evidence.While the translations follow the originals closely, the frequency differences are mostly in the direction of the implicit norm for the target language: English translations have fewer modals than Afrikaans originals, somewhat similar to English originals, and Afrikaans translations -where they deviate from the English originals -have more modals than the originals, somewhat similar to Afrikaans originals.This is most clearly seen in the period from 1935 to 1965.In 1925, the pattern for Afrikaans is distinct from the pattern established from 1935 onwards, which is consistent with the incomplete standardisation of Afrikaans at this point -the norms had not yet been established well enough for the translations to aim to approximate them.After 1965, the situation changes, and the translation frequencies deviate more from their respective originals, and move to much closer identity with the target language norms and deviate more from the source language frequencies.The data, as captured in Figure 3, pinpoint a change in respect of the translation of permission, possibility and ability modals from English to Afrikaans, and a change in the translation of volition and prediction modals from Afrikaans to English.The last two sampling points in the data are where bilingual contributions from individual speakers are at the highest, especially with Afrikaans speakers representing the National Party making an increasing proportion of the parliamentary contributions in English, which may disrupt the relatively stabler person-language relationships in the decades preceding 1975.
When benchmarked against the available findings for modals in Afrikaans and English in the 20th century more generally (see Table 2), it is clear that the parliamentary data contain much higher frequencies of modal auxiliaries than the written language in general.For South African English, Wasserman (2014: 226) reports a sharp decline in the overall frequency of modals and quasi-modals in South African English from the first half to the end of the 20th century.The Hansard data reveal that modals are consistently about 900 words per 100,000 more frequent for the periods that correspond to available data in Wasserman's (2014) study, just shy of 50% more frequent than in the written language generally.
For Afrikaans, Erasmus (2019: 199) reports a strong rise in the frequency of modals from the 1910s to the 1940s, and thereafter a more gradual decline to the 1970s and 2000s.8General written Afrikaans, as represented in Erasmus's data, shows only slightly higher frequencies of modals than general written English, as represented in Wasserman's data.By contrast, in the Hansard, Afrikaans modals occur at a frequency of more than 1700 words per 100,000, or almost 70% more frequent than in the general written corpus for the 20th century.
This overall comparison allows for the identification of a number of trends relevant to the research questions of this article.First, it is clear that (as might be anticipated from the functional context of parliamentary debates) modals are a particularly useful pragmatic resource, as indexed by the overall higher frequencies of modals in both languages in the parliamentary data, as compared to general-language corpora.Second, there is a contrastive difference between the two languages in that modals are in general language slightly more frequent in Afrikaans than in English; in the parliamentary context, this difference is exaggerated.Third, in the period under analysis, changes in the use of modals generally are also reflected in their use in parliament; however, variability is evident, which may be linked to the particular political configuration and changes in communicative dynamics in the years in question.Fourth, the overall pattern provides little evidence of a consistent convergence in the use of modal auxiliaries between the two languages in the bilingual parliamentary context over the 60 years in question; rather there are periods of convergent and divergent tendencies.The question of whether convergence occurs is taken up in more detail in the remainder of the discussion.Last, as far as translation is concerned, there is evidence of a notable shining-through effect, but some adherence to the overall quantitative norms of the target language during translation, in that deviations from similarity between translations and their source texts are in the direction of the norms of the target language.Deviations from this pattern appear to be motivated by social factors.

Modals by Semantic Groups
The overall picture discussed in Section 5.1 masks considerable variability among modal groups -some of which may, in the parliamentary context, be of more pragmatic value than others.Looking at specific modals in South African English, Wasserman (2014) reports that ought and shall decline to frequencies of close to zero, while the historical preterite forms should, would and might also decline significantly.The modal could shows a significant decline from the first half of the 20th century to the end of the century, but the decline is not significant compared to its frequency in the 19th century.By contrast, the modal can shows a significant increase throughout the 19th and 20th century, while will, must and may fluctuate without showing a long-term trend towards increase or decrease (Wasserman, 2014: 241).
In Afrikaans, the preterite forms sou and wou are already at a low frequency at the start of the 20th century, and by the beginning of the 21st century, have declined even further.The form mog has disappeared completely by the mid-20th century, while only kon remains stable in frequency throughout the 20th century.Among the non-preterite forms, mag fluctuates at a very low frequency (between 45 and 78 per 100,000 words throughout the 20th century), while sal, moet and wil likewise fluctuate, but at much higher frequencies.The modal that goes in the opposite direction over time is kan, which shows a strong increase from 353 to 514 per 100,000 words from the 1910s to the 2000s (Erasmus, 2019).
The combined frequencies of the three semantic groups, obligation and necessity (OBL), volition and prediction (VOL) and permission, possibility and ability (PERM) are presented in Figure 2, focussing only on the originals of the two languages first, while the corresponding numbers from the general language corpora are reported in Table 2.It becomes clear from Figure 2 that the main reason for the higher overall modal frequency in the original Afrikaans Hansard is due to the volition and prediction group, with the Afrikaans modals remaining between 100-400 examples per 100,000 words more frequent than their English counterparts.In the original Afrikaans parliamentary data the frequency of the volition and prediction modals increases quite sharply in 1935 and 1945 relative to 1925, but then declines sharply to a similar level as 1925 by 1955, with very gradual decline from 1965 to 1985.English volition and prediction modals start just above 1300 per 100,000 words in 1925, and their frequency remains relatively stable until 1965, when it declines more strongly.Both languages show a lower terminal point in 1985 than their starting point in 1925, and from 1955, the two languages remain at a relatively stable distance of around 200 modals per 100,000 words.These numbers can be benchmarked against the corresponding numbers from Wasserman (2014) and Erasmus (2019), which are summarised in Table 2.Not all semi-modals included by Wasserman (2014) are included here (only those that are included in our analysis of the Hansard data), whereas all the modals from Erasmus (2019) are included, as well as the modal behoort 'should/ought' .
The other two modal classes (permission, possibility and ability, and obligation and necessity) show strong fluctuations in the original Hansard data throughout the period.Both groups in Afrikaans show noticeable peaks in 1935 and 1965, the former of which also corresponds to a peak in the volition and prediction group in English.Except for the English permission, possibility and ability group, the other English and both Afrikaans modal groups end at a higher level in 1985 than where they started in 1925.However, compared to the volition and prediction group, the differences between original Afrikaans and English are relatively slight.
In comparison to the general historical corpora for the two languages, the original parliamentary Hansard data are at slightly higher levels only for the permission, possibility and ability group.The numbers for the obligation and necessity group are very different, though, with the Hansard data showing a much higher frequency than the general language.In the general language corpora, Afrikaans and English show a general decline in the frequency of these modals from the earlier to the latter half of the 20th century, not matched by a similar decline in the Hansard data.Parliamentary discourse quite clearly invites the expression of obligation, as illustrated by Example (1) and (2).9 (1) Ons moet almal toesien dat die boer nie van sy grond gedryf word nie.(Afrikaans original, 30 March 1965) 'We must all ensure that the farmer is not driven from his land' .(Hansard translation)10 (2) We have to protect the servant against the employer … (English original, 29 January 1935) A number of trends emerge from the general quantitative comparison of Afrikaans and English data in the Hansard and the general language, when considered by semantic groups of modals.The parliamentary Hansard makes much more extensive use of modals than the general language.This difference is due to the much higher frequency of modals of volition and prediction, as well as modals of obligation and necessity, while modals of permission, possibility and ability are used in similar frequencies in parliamentary and general language.At the aggregate level of semantic groups, clear evidence of convergent usage over time is not easily discernible; however, what is clear is that in the contact situation of the bilingual parliament the two languages are used in ways much more similar to each other than to the general language at the level of overall frequencies.This, in itself, provides some support for the suggestion that the direct contact situation of the bilingual parliament narrows the contrastive differences between the two languages.
To complete the picture, the translations are compared to their originals in Figure 3.The graph represents a subtraction of the normalised frequency of the modals per semantic group in translation from the corresponding normalised frequency in the originals.A number close to zero indicates quantitative similarity of originals and their translations, as is best demonstrated by the modals from the obligation and necessity group, especially from 1935 to 1965.A positive value indicates that the originals contain more modals of the semantic class than the translations, as illustrated by the volition and prediction modals of Afrikaans compared to their English translations, and the permission, possibility and ability modals of English, compared to their Afrikaans translations from 1935 to 1965.Negative values, most clearly illustrated by the translation of volition and prediction modals from English to Afrikaans, indicate that the translations contain more modals of the semantic class than the originals.
The most noticeable contrast is the large difference in respect of the volition and prediction category, where Afrikaans originals and translations clearly contain many more instances than the English translations or originals.A closer  inspection of the frequency of individual modals in the originals and translations shows that it is a straightforward case of higher frequency of the two Afrikaans forms sal (with mainly prediction or future meaning) and wil (with mainly personal volition meaning) being consistently used more in the originals and the translations than their English counterparts -will for future and would or WANT to for volition.Afrikaans volitional modal uses are therefore left untranslated in the English Hansard, as shown in Example (3), and English forms without volitional modals are translated with modals in Afrikaans, as shown in Example (4).The pattern of moderate difference, around 5-10%, is reasonably consistent from 1925 to 1975, but the difference increases very strongly in 1985 for the pair of Afrikaans originals and English translations, due mainly to a sharp decrease in frequency of WANT to in the translations against a level frequency of wil in the source texts.
(3) … ons wil hom baie voorspoed en sukses toewens … (Afrikaans original, 30 May 1985) … we wish him every success … (Hansard translation; direct translation is 'we want to wish him much prosperity and success') (4) Be that as it may.Let me ask him whether he means that, if they come into power, everything is just going to change immediately.(English original, 30 May 1985) Hoe dit ook al sy, ek wil hom vra of hy bedoel dat, as hulle aan bewind sou kom, alles eenvoudig onmiddellik sal verander.(Hansard translation -the italicised part backtranslates as 'I want to ask him') The modals of permission, possibility and ability show slow incremental change over time as far as Afrikaans to English translation is concerned, but mainly within 5% of identity between the frequencies of originals and translations By 1975, the combined frequency of may and could drop, without a corresponding drop in the frequency of kan.Afrikaans modals other than kan are infrequent throughout the 20th century, and thus a mismatch in translation equivalents for the various English modals and semi-modals in this group results in some omission of the English modal when translating to Afrikaans.The modals of obligation and necessity reveal a fairly close equivalence in frequency between the originals and translations in both directions, with the Downloaded from Brill.com12/02/2022 10:13:32AM via free access most notable exception being 1925, where English translations of Afrikaans originals use about 20% more modals than the originals.After 1925, the numbers stay relatively even, until a less extensive turn in the same direction in 1985.The Afrikaans expression of obligation and necessity is almost entirely performed by moet, with very infrequent use of moes and behoort.In English translations, should is the most frequent choice throughout, and at a lower frequency level, must and HAVE to remain fairly close to each other throughout.The biggest exception, which is precisely what contributes to the difference in 1925, is the much more extensive use of must as translation equivalent than in later -must is in fact used slightly more frequently than should itself.This should be attributed in part to direct translation of a somewhat less idiomatic kind, as illustrated by Example ( 6), where the use of both must and worry are dubious choices; "nie … moet lastig val nie" would better have been translated as "should not bother".
(6) Die rondskrywe was dat waar 'n [person van inheemse afkoms]11 besig is by 'n boer om te werk vir geld om sy opgaaf te betaal, die poliesie hom nie so veel moet lastig val nie.(Afrikaans original, 27 February 1925) The circular was to the effect that in places where a [person of indigenous descent] was employed in working for a farmer for money to pay his tax, the police must not worry him too much … (Hansard translation) Furthermore, should is often used not to translate a modal of obligation and necessity, but used in politeness formulae in English, such as I should like to ask with no corresponding obligation and necessity modal in the original Afrikaans, leading to a quantitative increase in the frequency of the modal group in the translations.Such uses are explored in more detail in Section 5.3.Translations follow the source texts quite closely, but due to non-identity of even cognate modals in English and Afrikaans, complete identity is not to be expected.The analysis of the data show that some early uncertainty in respect of the norms of Afrikaans, and the novelty of translation between English and Afrikaans lead to larger differences in 1925 than in subsequent years.Towards the end of the period, in 1975 and 1985, differences between translations and originals also increase, possibly due to the stronger force of the established conventions of the two target languages, and thus less shining-through and more alignment with target language norms.Some of the disruption may also be due to the increased practice of bilingual contribution by parliamentarians, particularly National Party parliamentarians who were native speakers of Afrikaans, but spoke in English more often than in the middle years.The English data, 11 Offensive word redacted.
and hence Afrikaans translations, were therefore confronted with new usage patterns deviating from established conventions, and new solutions had to be found.

Detailed Analysis: Modals of Volition and Prediction
The modals of volition and prediction reveal two corresponding trends in original Afrikaans and English: the gradual decline of the English modal will and its Afrikaans equivalent sal, and the increase in frequency of WANT to and wil (see Figure 4 and 5).Neither of these trends are matched by similar changes in the data for general 20th century written English (Wasserman, 2014) and Afrikaans (Erasmus, 2019), so possible explanations should be sought in the parliamentary data and the political and communicative context.This is the focus of the following discussion, which narrows the dataset to original Afrikaans and English only.
The modals sal and will are the most frequent modal in the respective languages, and convey future reference in the most unmarked way.They are also each other's closest translation equivalents, and the translated versions of the Hansard match the decline of the originals closely.To investigate whether their gradual frequency decline is associated with distinct usage patterns related to particular communicative contexts or aims, we examined the subjects and main verbs that combine with them.Sal and will combine with a wide range of syntactic subjects in both languages, with the impersonal third person singular pronoun it and its Afrikaans equivalent dit being the single most frequent one, alongside a very diverse range of other third person subjects.The epistemic meaning of prediction is very prominent among the uses of sal and will, as illustrated by Example ( 7) and ( 8).Over time, the extent to which the prediction meaning is used by the speakers declines.This is not offset by any other emerging uses that increase in frequency.However, the declining lines also show some fluctuation, which can be accounted for in part by the second very prominent pattern in the use of will and sal, which is their combination with the noun Minister as subject.This noun is the most frequent non-pronoun subject combining with these two modals.Such uses occur either in direct questions posed to the minister, illustrated by Example (9), or as complement clauses after a complement-taking verb with the speaker as subject, illustrated by Example (10).In both cases, The frequency of these combinations fluctuates extensively over time, as shown in Figure 6.Afrikaans usage picks up from extremely infrequent in 1925 to very frequent in 1935 and 1945, after which the frequency declines dramatically again.English usage is consistently somewhat more frequent than Afrikaans usage, but shows a persistent increase after 1945, to peak in 1965, after declining to similar levels than Afrikaans.These two local peaks in 1945 for Afrikaans and 1965 for English correspond to outliers of similar absolute magnitude in Figure 4 and 5 for these two languages respectively in the overall frequency trajectory of sal and will respectively.
A close reading of the data shows that these changing fortunes are not due to linguistic differences between Afrikaans and English: the actual patterns of use are quite similar throughout, with structures like that in Example (9) more frequent than structures like that in Example (10) for both languages.Rather, the use of such a line of questioning with a minister is typical of opposition politicians.When the Afrikaans National Party form the opposition in 1935 and 1945, they use this resource very frequently, as shown in Example (9).By contrast, when the mainly English United Party is in opposition after losing the 1948 election, they find probing questions to the minister more advantageous to their rhetorical strategy, a strategy continued to some degree by the Progressive Federal Party as they gradually supplant the United Party as opposition from 1975 onwards, as shown in Example (10).
The second important development evident in Figure 4 and 5 is the rise in frequency of the modal that conveys the deontic sense of volition most directly, wil in Afrikaans and WANT to in English.This pattern of change is matched by a corresponding change in the translations too.In Afrikaans, the frequency increase is such that wil became more frequent than sal in Afrikaans from 1975, and WANT to drew level with would as second most frequent modal in English at the same time.The data reveal that the first person singular pronoun I/ek is the subject of clauses containing the modal WANT to/wil safely more half the time (immediate left or right collocate in Afrikaans 2789/4449, and immediate left collocate in English 709/1288), and no other subject has a tenth of the frequency of the first person singular pronoun (second most frequent subject in Afrikaans is ons 'we' at 209/4449, and in English we 68/1288).
The expressions I want to/ek wil are used in illocutionary formulas, with main verbs say and ask (Afrikaans sê and vra) as the two most frequent verbs used (illustrated by Example ( 11) and ( 12)), alongside other performative verbs with more specific verbal act meanings.In Figure 7a and 7b, normalised frequencies for the subject ek/I and the verb sê/say are reported.In both languages, the use of the performative template without any modal (ek sê/I say) is initially the most frequent use, and remains so until 1955 for Afrikaans and 1965 for English.English even shows an increase in this usage until 1945 (when the United Party was in power), while Afrikaans shows a slight increase in 1945/1955 (on both sides of the transition to power from the United Party to the National Party).These rises are offset by slight but corresponding decreases in the use of all other modals together (e.g.ek moet sê/I must say, or in English also I would (like to) say) until 1965.From 1965, however, the use of wil in Afrikaans and WANT to in English rises quite sharply, as if there is a simultaneous congruent change in the preferred manner to introduce statements by speakers of both languages and on both sides of the aisle.
There is a levelling off and slight decline in 1985 again, and the overall trajectory of the formula ek wil sê/I want to say matches the changed trajectory of wil/WANT to in the overall frequencies (Figure 4 and 5) quite closely.
In Figure 8a and 8b, the normalised frequencies of ek vra/I ask with or without a modal are reported.Once again, the fortunes of combinations with wil/WANT to are important, but the overall patterns are less clear-cut for the remainder of the options.Frequency increases for ek wil vra/I want to ask correspond to the years in which speakers of the two languages were in the opposition, matching the pattern detected earlier for combinations with die Minister sal/the Minister will (Figure 6).In Afrikaans, other modals occur at negligible frequencies in this combination, but the combination without a modal, e.g.Example (13), remains an option throughout, and even rises towards the end of the period.The option without a modal is quite direct and impolite, and seems to be particularly useful for very critical questions, or to ask rhetorical questions.
( The analysis of the usage patterns of volition and prediction modals reveals a number of strong similarities between Afrikaans and English in the parliamentary context.Similar long-term frequency changes are identified, which do not reflect long-term changes in the general language.In addition, certain Afrikaans and English structures are harnessed for their pragmatic value as illocutionary formulas.A subset of these formulas also reveal themselves as particularly useful to politicians in the parliamentary opposition, and thus what manifests as different frequency developments between English and Afrikaans are actually reflexes of changes in the parliamentary control, with politicians using Afrikaans and English respectively doing similar things, at different times, when they find themselves in the opposition.

Conclusion
This article set out to answer three research questions in connection with the use of modal auxiliaries in the language of the bilingual South African parliament of 1910-1994, within a more general attempt to reflect on the methodological challenge within contrastive pragmatics of finding data to compare on the basis of sufficient pragmatic or functional similarity to render the comparison meaningful.
The first question concerns contrastive differences in the use of modals.The data reveal that Afrikaans modals are more frequent than English modals, but there is a larger similarity in that the parliamentary data, irrespective of language, contain considerably higher frequencies of modals than general writing.Looking beyond the individual modals, which correspond to varying degrees between the two languages, the three semantic groupings of modals reveal broadly similar frequencies and uses.
The second question concerns pragmatic change and the possibility of convergence.At a quantitative level, changes do not reveal convergence, although there are broad patterns of similarity in the direction of change of the most frequent modals, such as English will and can and Afrikaans sal and kan.In addition, in the contact situation of the bilingual parliament the two languages are used in ways that are more similar to each other than to the general language at the level of overall frequencies.This does suggest that the direct contact situation of the bilingual parliament reduces the contrastive differences between the two languages.
At a more specific level, a number of illocutionary formulas were identified, used for managing politeness and ways of addressing politicians from across the aisle.These formulas are cross-linguistically very similar at a functional level, and some are used most often by politicians in the role of the parliamentary opposition.Thus, functional similarities between the two languages do not present themselves in quantitative terms, because the languages are associated more or less strongly with opposition roles at different times over the course of the 20th century.
The third question concerns the way translations reflect the pragmatic differences in the use of modals in original Afrikaans and English respectively.The data reveal that translations are strongly responsive to the original texts, which makes sense in the broad legal framework of the Hansard, where fidelity is an essential quality in official record-keeping.Nevertheless, the shining-through of the properties of the original texts is ameliorated by adjustments in the direction of the statistical norms of the target language, such that English translations contain fewer modals than Afrikaans originals, and Afrikaans translations contain more modals than English originals.The analysis of translations reveals instability of the norms at the start of the process of bilingual record-keeping in 1925, before settling in stable patterns throughout the middle part of the twentieth century, with renewed instability towards the end of the latter years of the bilingual parliament, as Afrikaans speakers came to use both languages more often, thereby introducing new variability in the original texts, which in turn affect the translations too.Despite these points of instability, the overall relationship is quite clear: translations do not lead the changes, but reflect them, confirming the finding of Redelinghuys (2019) with a narrower dataset but better temporal resolution and evenness of distribution of sampling points.
The methodological challenge in contrastive pragmatics that this article attempted to address is the notion of sufficient similarity of contexts for comparison.A number of overarching similarities between the parliamentary data and the general patterns of use of modals, as described by Wasserman (2014) and Erasmus (2019), have been identified.Unique features of parliamentary language, shared by the Afrikaans and English data, have also emerged, reinforcing the importance of the context of comparison for valid findings.However, most strikingly, quantitative deviations over time turn out to be functional similarities that present themselves in different languages at different times, due to the political roles (as governing or opposition party) that speakers of the two languages played at different times.Without access to the level of context and granularity of analysis, the presumed differences between the languages may have been overstated and left unexplained.

Figure 2
Figure 2 Normalised frequency (per 100,000 words) of modals expressing obligation and necessity (OBL), volition and prediction (VOL) and permission, possibility and ability (PERM), in the original Afrikaans (A) and English (E) Hansard material

Figure 3
Figure 3 Percentage of change in normalised frequency of modals in translation for modals expressing obligation and necessity (OBL), volition and prediction (VOL) and permission, possibility and ability (PERM), from originals in Afrikaans (Ao) to translations in English (Etr) and originals in English (Eo) to translations in Afrikaans (Atr) . The translation of English to Afrikaans shows more extensive change, being close to identity in 1925 and returning to this level in 1975 and 1985, but with English originals showing approximately 10% more modals than the Afrikaans translations from 1935-1965.Considering the individual modals, the 1925 data show more extensive use of Afrikaans mag, as translation equivalent of the even more frequent may.Thereafter, mag declines to half its 1925 frequency and stays there, without a corresponding decline in frequency of may before 1975.The rest of the deviation is due to the fluctuation in the selection of kan, the most frequent translation equivalent (and most frequent permission, possibility and ability modal of all) -which often translates to may and could, alongside can.The data indicate a stabilisation of Afrikaans target language norms after 1925, where mag has a much narrower range of uses than its English cognate may, rather than too direct translation of English in 1925, illustrated by Example (5).The Afrikaans translation is quite direct, but less idiomatic, at least by present-day standards, and there are no corresponding uses in the original Afrikaans records, while the Afrikaans translations reveal a small number of further similar examples, all from 1925 except for one from 1945.(5) It may be said that as I do not represent a constituency in the Transvaal this matter does not concern me … (English original, 27 February 1925) Dit mag gesê word dat hierdie saak my nie aangaan nie, daar ek nie 'n kiesafdeling in Transvaal verteenwoordig nie … (Hansard translation)

Figure 4
Figure 4 Normalised frequency (per 100,000 words) of Afrikaans modals expressing volition and prediction (original Afrikaans)

Figure 5 Figure 6
Figure 5 Normalised frequency (per 100,000 words) of English modals expressing volition and prediction (original English)

( 11 )
Ek wil net sê dat gedurende hierdie Begrotingsdebat die Opposisie hoegenaamd niks opbouends aan die hand gegee het nie.(Afrikaans original, 31 March 1965) 'I just want to say that during this Budget debate, the Opposition have made no constructive suggestions whatsoever' .(Hansard translation)(12) I want to ask a few questions and I do hope that some answers will be forthcoming.(Englishoriginal, 29 May 1975)

Figure 8a
Figure 8a Normalised frequency (per 100,000 words) of the combination ek + (volition/ prediction modal) + vra in Afrikaans original material

Figure 8b
Figure 8b Normalised frequency (per 100,000 words) of the combination I + (volition/ prediction modal) + ask in English original material should, must, HAVE to and ought to in English,  and moet 'must, HAVE to' , moes 'had to' and behoort 'ought to' in Afrikaans.2.Permission, possibility and ability: can, could, may, might and BE ableto in English, and kan 'can, BE able to', kon 'could', mag 'may', and mog 'might' in Afrikaans.3. Volition and prediction: will, would, shall, BE going to and WANT to in English, and sal 'will, shall' , sou 'would, should' , wil 'WANT to' and wou 'wanted to' in Afrikaans.Data analysis was done by means of frequency comparisons and collocation analyses in WordSmith Tools.
the English original texts mostly have fewer modal auxiliaries than the other datasets, although the Afrikaans translations are very close to the English originals from 1935 to 1965.The English translations generally follow the trajectory of increase and decrease of the Afrikaans originals at a slightly lower frequency, although the frequency difference increases over time.

Afrikaans (adjusted from Erasmus, 2019) English (from Wasserman, 2014) 1910s and 1940s 1970s and 2000s First half of 20th century 1990s
13) Ek vra die Minister want dit blyk nie uit die begrotingsrede.(Afrikaansoriginal,28February1935)'Iaskthe Minister because it does not appear from the Budget speech' .(Hansardtranslation)InEnglish,othermodalstogetheraremore frequent early on, especially would, which remains more frequent than WANT to until 1955, before being surpassed very clearly by WANT to.The modal would often expands into the even more polite collocation would like to, as illustrated by Example (14), with the alternative should like to, as illustrated by Example (15), overtaking would like to towards the end, and contributing to the overall frequency of the obligation and necessity group, despite being employed functionally in competition with the volition and prediction group.Also towards the end of the period, culminating in 1985, the use of may becomes a viable competitor, especially when used to ask a direct question, as shown in Example (16).The English formulations van Rooy and Kotze Contrastive PragmaticS 3 (2022)show more frequent use of politeness markers than the Afrikaans formulations on average, and also show a much wider range of options than Afrikaans.Nonetheless, would almost disappears and WANT to asserts itself as dominant option, which increases the degree of correspondence between Afrikaans and English.(14) I would like to ask the hon.Minister in all seriousness, how can the machinery of this Act prevent such a situation?(English original, 30 May 1955) (15) I should like to ask the hon. the Minister why he feels that in this instance it is necessary to go further than that.(English original, 29 May 1975) (16) Mr Speaker, may I ask the hon the Deputy Minister why he did not amend the Bill along the lines suggested by the HSRC?(English original, 28 February 1985)