Two Recent Major Afrikaans – English / English – Afrikaans Dictionaries from Pharos

When Pharos Dictionaries was established in 1996, its first order of business was to develop a comprehensive Afrikaans–English/English–Afrikaans dictionary that could succeed the standard-bearing but ageing TW (Tweetalige Woordeboek/Bilingual Dictionary by Bosman, Van der Merwe and Hiemstra). The article explains in detail and with examples how this new dictionary, Pharos One for short, was developed from the TW. Information about syllabification, grammar and usage were greatly enhanced, the lemma structure was made more accessible and thousands of new words and expressions were added. The most difficult aspect of the development process was the incorporation of older material from the TW. For the sake of comprehensiveness much of this material was retained, but because of the sensitive nature of many words and expressions from the old South Africa, usage instructions had to be added with great care. For the editing of the English– Afrikaans half the Pharos editors could draw on a plethora of quality British and American and bilingual European dictionaries, but for the Afrikaans–English half the editors had to rely on their own language knowledge. The article explains how with reference to electronic archives and the help of outside consultants the Pharos team could verify their judgment. In response to lagging sales, Pharos embarks on the development of a scaled-down version of Pharos One that would retain its best characteristics, but would have a more manageable size and affordable price. The article explains how a one-third reduction of Pharos One was achieved to produce this new dictionary, called the Concise for short.


1.
A new South African dictionary publisher Pharos Dictionaries was founded in 1996 to amalgamate the various dictionary titles from Naspers.Dictionaries hitherto published by Tafelberg and J.L van Schaik, for example, would from now on bear the Pharos imprint.The company would also provide a home for those editors with a lexicographical bent to ply their trade.With manager Hans Büttner, and with editors Madaleine du Plessis and Jana Luther to tackle the text work, Pharos Dictionaries set about modernising the titles it had inherited, but also preparing for the publication of brand new titles.
A thorough revision of Bosman, Van der Merwe and Hiemstra's Tweetalige Woordeboek/Bilingual Dictionary (the TW) was the fledgling company's most immediate and pressing task.Generally considered the standard, authoritative reference work for translating English to Afrikaans and vice versa, the TW was increasingly falling behind the requirements of language practitioners in particular and, in general, anyone who regularly had to write clearly, correctly and idiomatically in both Afrikaans and English.The last revision of the TW, the eighth (TW8), by Joubert and Spies, was completed in 1984 and with the establishment of Pharos Dictionaries that was already 12 years ago.In dictionary terms, that is a fairly lengthy period.Large overseas dictionary publishers like Oxford and Collins seem to manage at least two revisions of their frontline titles every decade.
A further complicating factor was that the years that had lapsed since the publication of TW8 were no ordinary years.During this period the Berlin Wall had fallen and here at home South Africans had their first fully democratic elections.Both these events brought about a complete reordering of the social, political and geographical structure, globally and locally.On the technological front these were equally revolutionary times.Personal computers became commonplace, and the Internet and cell phones opened up a whole new world.It was also the era in which the shadow of Aids began to creep across the globe and phenomena like the greenhouse effect began to make their presence felt.Tumultuous developments such as these find their expression in language almost simultaneously, and within a reasonable time dictionary users expect to find in their favourite dictionaries the new vocabulary generated in this way.
Work on the revision of TW8 started in earnest soon after the establishment of Pharos Dictionaries.Madaleine du Plessis was chief editor, and two more editors were appointed for the task, namely Wanda Smith-Müller and Fred Pheiffer.Although the revision of TW8 would consume most of Pharos's time and energy over the next couple of years, this was by no means its only project.Its first two publications were both aimed at meeting the demand for an up-to-date Afrikaans-English/English-Afrikaans (AEEA) dictionary.The 14th edition of the Groot Woordeboek/Major Dictionary edited by Louis Eksteen, a title inherited from J.L. van Schaik, provided some improvement on TW8, but was not sufficiently modernised to replace it and ultimately fell short of expectations.The Tweetalige Frasewoordeboek/Bilingual Phrase Dictionary by Phillip Joubert, a title inherited from Tafelberg, proved to be very useful, but as a specialised dictionary it was intended at best to augment the TW.
Pharos's first 'home-grown' publication was New Words/Nuwe Woorde by Madaleine du Plessis in 1999.It contained Afrikaans translations for well over 10 000 modern English words and expressions, material hitherto not covered by the TW, but also new, additional meanings and uses of terms already contained in the TW.This publication was the first product of the revision work being done on TW8 and as such provided a sample of what the successor to the TW would offer.It contained words used in a wide range of disciplines and words reflecting technological advances, but also many new phrases and idioms and a good number of informal and slang words and expressions.However, as Joubert's Phrase Dictionary, it was designed to be used as a supplement to the TW, not as a standalone dictionary.Although all three these dictionaries were ultimately superseded by or assimilated into the new AEEA dictionary, they all proved their worth as vital stand-ins during the gap between the TW and its successor, a period that eventually stretched to two decades.Moreover, they became essential cogs in Pharos's first electronic product, the Pharos 5 in 1 Dictionaries/Woordeboeke CD-ROM, a vehicle that prolonged their usefulness.It was first released in 2000 and is still proving popular in 2008.

2.
A successor to the TW The development of a new top-of-the-range AEEA dictionary to succeed the TW encompassed the following additions and alterations to the text of TW8: inserting syllable separators and underlining the stressed syllable in all headwords and their derivatives; restructuring many of the entries; shuffling the order of the translations to place current meanings first, and older or specialised meanings further back; adding new material; and treating dated material.

Syllable treatment
Below is a comparative entry from TW8 and Pharos One, showing the addition of syllable separator dots and stress underlining in both the main entry and the derivatives in the latter: Note also the expansion of the grammatical and usage information in Pharos One.
Syllabification was generally easier to effect on the Afrikaans-English (AE) half than the English-Afrikaans (EA) half of the dictionary, since Afrikaans syllabification is phonetic, and to apply it is relatively straightforward for anyone with a sound knowledge of spoken Afrikaans.English syllabification on the other hand is morphological and requires knowledge of the underlying rules or, alternatively, reference to sources with English syllabification.Surprisingly, British English dictionaries generally do not contain syllabification.The 1985 edition of the Collins Dictionary of the English Language seems to have been the last to offer comprehensive information on it.Bilingual European dictionaries with an English component like those of Van Dale and Langenscheidt show syllabification and American English dictionaries like the American Heritage and the online versions of Merriam-Webster and Encarta all have syllabified lemmas.
Intriguingly, the syllabifications offered in these sources do not always match.The table below shows the differing treatment of a few randomly picked examples from the above-mentioned edition of Collins and the 1993 edition of The American Heritage College Dictionary.The difference in interpretation seems to be influenced by variation in the placement of the stressed syllable (for example, both homophyly and homophyly are acceptable) but the two sources do not offer different syllabifications for the variant pronunciations.

Collins American Heritage
Admittedly, these are all fairly low-frequency technical words, except maybe for demonstrator and disinfectant, and the level of agreement in the above-mentioned sources is probably close to 99%, but in the quest for preciseness in one's own dictionary discrepancies in the sources one relies on can be frustrating.

Restructuring
Many of the entries in TW8 tend to be very long and overly complex, because of an attempt to place as many derivatives and compounds under a single lemma as possible.This type of lemma organisation may be highly systematic, but it is not necessarily user-friendly.In an effort to make the information in the entries as accessible as possible, such long entries had to be disentangled.In particular, wherever derivatives of the main entry were placed among compounds of the same entry, they were removed and elevated to separate main entries.Hence derivatives and compounds were not allowed to coexist in the same main entry (except of course where the derivatives followed after the compounds, alphabetically speaking).Below are comparative entries from TW8 and Pharos One; in the latter derivatives of ego have been made separate entries.The main advantage is that the reference of the compounds' tilde to the main entry is apparent, without interference from derived forms.Note also the expanded use of usage labels and example phrases in the Pharos One excerpt.
Performing such disentanglement procedures manually (at this stage of the editorial process cutting and pasting in XYWrite, an obsolete yet very stable word-processing program) posed at least two risks.Firstly, derivatives could be removed from an entry and posted as new entries immediately after their mother entry without checking whether this procedure disturbed the alphabetical order of the entries that previously followed immediately after the mother entry.A second risk is that in the cutting and pasting of derivatives one could get distracted and forget to paste.In TW8 the lemma suit (p.1235) has the derivatives suitability, suitable, suitably, suited, suiting and suitor dispersed among the compounds suitcase, suitcase cover and suit length, but in Pharos One only suited and suiting survived (p.1407), which is most likely an oversight of the kind described above.The first risk can be obviated by regularly running a script over the lemma list to check for alphabetical errors, but the second risk is more serious.A dedicated lexicographical program like TshwaneLex, which Pharos Dictionaries started using in 2006, probably offers the best solution.Not only does it alphabetise automatically, but also alerts the user whenever deletion of any part of a lemma is attempted.

Reordering translations
Any revision or renewal of the semantic content of the lemmas in TW8 relied in the first instance on the lexicographer's own sense of language, on a hunch, admittedly, that the translations given in TW8 did not quite match the current usage of Afrikaans or English in South Africa.Although constantly aware that such an approach was perhaps not best practice, the constraints of time left one with very little other option.Fortunately, with the revision of the EA half of TW8, which was tackled first, one had the luxury of being able to refer to several first-class, up-to-date English dictionaries.In cases where one felt TW8 had become dated or deficient, one could compare its translations with the meanings offered by Collins or Oxford, for example.Such a comparison showed that TW8's content was basically sound, but confirmed that it had dated in several respects.Below are two examples comparing entries from TW8 and Pharos One, which show how the focus of the meaning of the words concerned has shifted.
In the first example the translation of reduction by 'terugbrenging', which is unlikely to find much resonance nowadays, has been pushed right back in Pharos One, allowing the expected translation of 'vermindering' etc. to stand first.Note the use of the label (arg.) to indicate that certain translations have become archaic.Further note how Pharos One places usage labels like (wisk.) and (med.)before a translation, where they can refer to both the lemma and its translation.In the second example, the usual translation of uitskel one would nowadays expect, namely 'scold', is buried deep in the TW8 entry.In the Pharos One entry, 'scold' has been elevated to the foremost position and 'abuse' has been modified by 'verbally'.The example phrase iemand uitskel has been extended considerably with idiomatic translations, and the name-calling aspect has been relegated to the back of queue.Note how the scope of the second example phrase has been broadened in Pharos One by the use of the ellipsis and the offering of an example (liar etc.) as a possible complement.Compared to the EA half, the editing of the AE half had to proceed without fresh Afrikaans lexicographical resources.Pharos's own Verklarende Afrikaanse Woordeboek (the VAW) predated 1994 -that benchmark date for all modern South African reference works.A new edition of the Verklarende Handwoordeboek van die Afrikaanse Taal (the HAT) was published in 2000, but it was not vastly different from earlier editions.Of the standard reference work of the Afrikaans language, the Woordeboek van die Afrikaanse Taal (the WAT), less than two-thirds of the letters had been done (the letters N to O were completed during the editing period of Pharos One).Moreover, many of the WAT's earlier volumes had become severely outdated and as a result contained the very flaws that one sought solutions for in order to produce Pharos One.
The implication of this paucity of suitable references was that one had to rely -even more so than in the EA half -on one's own knowledge of language, in particular of Afrikaans in its present-day form.Pharos did have at its disposal two types of corrective or control though.The first of these are the Media24 newspaper and magazine archives and the archive Pharos has built up from manuscripts from its sister companies like Tafelberg and Human & Rousseau and from other material.These electronic databases provide a means of checking the frequency, usage and most current meaning of words and expressions.And although gleaning information in this way is probably the ideal approach to lexicography, working with raw data is a laborious process.Because time was of the essence, this research method could only be resorted to when one's own ability to judge with certainty the validity or correctness of a TW8 entry failed.
The second corrective was to employ the services of a number of seasoned Afrikaans linguists (Fritz Ponelis, Ernst Kotzé, Tom McLachlan and the late Johan Combrink) and editors (Louise Voigt and George Boshoff) to read and comment on the Pharos One manuscript in the various stages of its preparation.As outsiders not intimately involved in the project, they could provide input that was objective and valued for its authoritativeness.

New material
The bulk of the new material in Pharos One came straight from New Words/Nuwe Woorde.It was added to the EA half of Pharos One (New Words being one-directional EA only) either as new lemmas or integrated into existing entries as new meanings, expressions, compounds or derivatives.In the example below, the top excerpt is the original section from TW8 and the middle one from New Words while the bottom excerpt shows the amalgamation in Pharos One.In the excerpt from New Words, Fosbury flop provides Pharos One with an altogether new entry, fossil fuel is a new compound to an existing noun entry, while fossilised is the adjectival use of the past participle of an existing verb.Note once again how the derivatives in the fossil entry in TW8 have been removed and elevated into independent entries in the Pharos One excerpt to allow the compounds of fossil to be closer to and follow uninterrupted after the lemma.Alphabetisation in the new fossilise entry has been suspended so that fossilisation can be listed as a derivative of the lemma.
Note too once again the expanded use of usage labels in the Pharos One excerpt.The approach throughout to apparently dated or rare material in TW8 was not to summarily dismiss such entries, but to try to accommodate as many of them as possible through the clever application of usage labels.Hence (arg.), (argeol.) and (Austr., infml.) to indicate archaic, archaeology and Australian English or informal.However, an example like foskorite, which has disappeared from Pharos One, would have been unknown to the editor working on that section of TW8, would furthermore not have been found in any of the British, American, Dutch or German dictionaries in the Pharos office, including a venerable edition of Webster's, and finally, would in all likelihood have returned a low Google search score (less than one hundred).
Where new material in the EA half of Pharos One gave rise to new Afrikaans translations too, a 'mirror image' of the entry was created allowing the translations to be used as new lemmas in the AE half.In the excerpts below, the new compounds of gift in New Words gave rise to the following new compounds of geskenk in Pharos One: geskenkpak(kie), geskenkdoos, geskenkbelasting and geskenkpapier.Other sources of new English material were The New Oxford Dictionary of English, particularly the Concise Oxford Dictionary, and the Collins English Dictionary.A relatively late source of new words was the seventh edition of Roberts Birds of Southern Africa.In this edition, a number of the local English bird names were changed to make them correspond with international English bird names (lapwing instead of plover, turaco instead of loerie, thick-knee instead of dikkop, etc.).As many of these changes as possible were implemented in Pharos One, but because the letters were typeset as they were completed, this was not always practical.Nevertheless, Pharos One was probably one of the first reference works to contain the new nomenclature from Roberts.

Dated material
As described in the previous section, the editing policy applied during the preparation of Pharos One was to preserve as much of the dated material as possible, rather than get rid of it wholesale, which would have been the kneejerk approach.However, preserving dated material for the sake of comprehensiveness implied a responsibility to the dictionary user to identify such material with care and to label it accordingly.Dated material was classified as obsolete (old-fashioned and no longer current), archaic (language relics used occasionally for effect), rare (not current, but not necessarily dated) or historic (the object referred to is dated, but not necessarily the word).Obviously these categories overlapped and labelling dated material one way or the other was not a straightforward task.However, as mentioned above, having established English dictionaries as references during the preparation of the EA half was a great boon.Especially the New Oxford and the Concise Oxford are profuse in their labelling.For the identification and labelling of dated material in the AE half, the same practice was followed as described in section 2.3.Given the history of racial oppression in South Africa, great care had to be taken to isolate terms and expressions that still reflected racism.Many blatantly racist words had already been labelled as derogatory in TW8 (see for example pp.235 and 964) and in Pharos One this process was continued with rigour.For example, for each of the myriad Afrikaans plant or animal names that is a compound of either hottentot or kaffer a non-offensive alternative was found (see pp. 230 and 256).Words such as Bantoe/Bantu (TW8, pp.46 and 710; Pharos One, pp.44 and 796), Kleurling/Coloured (TW8, pp.254 and 769; Pharos One, pp.276 and 865) and naturel/native (TW8, pp.338 and 1036; Pharos One, pp.368 and 1179) that were still deemed inoffensive in the days of the TW, or at least not marked in it as problematic, were labelled as derogatory and as obsolete or historic in Pharos One.
A certain group of Afrikaans idiomatic expressions proved quite vexing to categorise and label.Although hardly derogatory, their gleeful extolling of physical violence is offensive and certainly out of step with a culture of human rights.The English translations with their emphasis on goodness and soundness are not above reproach either.Leaving them unmarked would therefore have been insensitive, but how to deal with them?Judging from their variety, these expressions still seem very much part of the Afrikaans vocabulary.As to their frequency of use, the corpora contained in the Pharos's text archives would have thrown light on that but the urgency of the work on Pharos One precluded a thorough investigation.However, a cursory search of the newspaper archives indicated that their use tends towards the figurative and the colloquial, and that they are often employed in sport journalism, describing rugby matches for example.This is how some of these expressions were eventually labelled, i.e. (infml.)or (sport journ.),hardly satisfactory, but dictionary tags are quite limited when attempting to categorise such a subtle issue.In any case, it was only after the editing of the AE half was well on its way that the extent of these expressions became apparent and, as noted at the end of section 2.4, unless one had to deal with an error one tried to avoid altering pages that were already typeset.
Generally speaking though, adding new material to Pharos One was by far the easier operation than dealing with dated material.As the editing work on the text of TW8 progressed, a distinct impression began to take root that during the life of the TW, successive editions must have been augmented by new material, but not enough was done to evaluate existing material for its currency or appropriateness.It would therefore hardly be surprising if a study found that certain words and expressions ingrained in the TW survived from the first half of the previous century.All those Afrikaans idiomatic expressions steeped in an agrarian past are a case in point.Quaint they may have become, but do they really still speak to a modern, urbanised youth?How long can one justify retaining them in a dictionary, even for the sake of comprehensiveness?
With close to a decade of intense, dedicated and informed revision behind it, Pharos One can rightly claim to be the most authoritative and up-to-date AEEA dictionary by far.Yet what the issues raised above point to is that the dictionary still carries a lot of baggage from the past which has not been unpacked and sorted.For it to become truly modern and revitalised, its contents would have to be evaluated far more rigorously against text corpora than has been possible up to now.However, this is a labour-intensive and time-consuming approach, and whether the means that such an approach requires could readily be made available in commercial publishing remains a moot point.

The Concise
Relatively soon after the publication of Pharos One at the beginning of 2005, the marketing team of NB Publishers, of which Pharos Dictionaries is an imprint, began to report that the market was daunted by the size and price of Pharos One.With dimensions of 21 cm x 27 cm x 5 cm and a mass of 2,5 kg, it is quite a hefty book, and for many the initial price was probably on the steep side.After having been cloistered with Pharos One for so long, laboriously poring over the text, it was disappointing that one's enthusiasm at its release was not immediately and generally shared.Perhaps one had hoped, somewhat naïvely and vainly, that many of the people who owned a copy of TW8 (of which more than 170 000 copies were sold over its 20-year print run) would buy a copy of Pharos One the moment it was released.The initial sales of Pharos One were certainly not poor; in fact, they matched the average monthly sales of TW8 during the time it was a Pharos imprint.But that would be comparing Pharos One to the TW at the end of its long and useful life -the average monthly sales of TW8 over its entire 20-year print run are about double that of Pharos One.The 30 000-copy first print run of TW8 sold out in just two years.What seems to have been overlooked was that in the heyday of the TW, South Africa was officially and effectively a bilingual country.Every public office had to be able to conduct its business in both English and Afrikaans.In particular, every institution or company involved in the legal system -Parliament, the courts, academic legal faculties, legal companies -had to produce its documents in English and Afrikaans.This created a huge demand for a dictionary like the TW.Since 1994 however, and even somewhat before that, this requirement has largely fallen away.The fact that Afrikaans is no longer officially prescribed (except in provinces like the Western Cape where Afrikaans is a major language) by itself could explain the marked difference between sales of TW8 and Pharos One.
Although the initial sales of Pharos One were slower than anticipated, there was clearly a demand for a new top-of-the-range AEEA dictionary.In order to capitalise on this demand, and also to recoup the considerable investment made to deliver Pharos One, NB Publishers recommended to Pharos Dictionaries to produce a dictionary offering the same scope and quality as Pharos One, but of a more economic price and a more manageable size, a Pharos One Lite essentially.Crucially, Pharos was given just two years to accomplish this.A tall order certainly, but not impossible, since the secondary purpose for which Pharos One was designed was to serve as a database from which smaller dictionaries could be extracted -a mother dictionary that could generate offspring -especially after the text was contained in TshwaneLex.Unfortunately, by the time Pharos Dictionaries started evaluating TshwaneLex late in 2005, the new dictionary project, which would become the Concise, had already progressed too far for it to be hauled into this new program.The work schedule was just too tight for a conversion process that would undoubtedly have caused time-consuming teething problems.The Concise was done in MS Word from the start, certainly not ideal, since processes like alphabetisation and cross-referencing had to be done manually, but it sufficed.
Because each page of Pharos One is made up of three columns, cutting the equivalent of one column per page would bring about a dictionary two thirds the size of Pharos One.A cursory investigation showed there was enough 'nonessential' material on each page to make such a cut feasible.Clearly the comprehensiveness of Pharos One would have to be sacrificed, the dated material from the TW, treated with such deference in Pharos One, being the obvious first candidate for elimination.The care with which dated material was labelled in Pharos One made this process fairly easy.Two further groups of words were identified that could be cut in toto, namely subject or technical terminology, and typically British, American or Australian English.However, an important proviso was that if any word in these three groups had become assimilated into everyday Afrikaans or South African English, it would be retained.
The editorial team established for the Concise consisted of Pharos One editors Jana Luther and Wanda Smith-Müller, freelance editor Celia Slater and Fred Pheiffer as final editor.Its ability to judge what should be eliminated and what retained was of critical importance for the successful and timeous completion of the Concise project.The editorial working conditions for the Concise were essentially the same as those for Pharos One -quality English dictionaries for the EA half, own judgement and, where time permitted, reference to electronic archives for the AE half.
Apart from reducing each page of Pharos One by a third, three other processes were carried out: more new words were added, errors in Pharos One that the editorial team had become aware of, were rectified and, where desirable or possible, the Pharos One lemma structure was simplified.In most respects this represented a continuation and further refinement of the work done on Pharos One.

Reduction, new material and corrections
The table below compares a fairly randomly chosen page from Pharos One with the corresponding section in the Concise.Of the 82 Pharos One lemmas, 32 (or 39%) were removed completely, while 14 had parts like example phrases, compounds or derivatives deleted.As can be seen from the Pharos One labels, the lemmas that have been removed are mostly dated, rare, technical or British or American peculiarities.
Note the new Concise entry neomycin.The Concise contains at least 1 000 new words, mostly in the EA half, with some, through translation, giving rise to new Afrikaans entries as well.Many of the new English entries were sourced from recent English dictionaries, in particular the South African Concise Oxford Dictionary.
In the above section from Pharos One, a mistake had crept in: among the compounds of net 1 is network, given with translations and an example phrase.However, further on, network occurs again in its own fairly extensive entry, with a verb homograph, several more example phrases, and two derivatives, networker and networking.In the Concise, this mistake has been rectified: network is retained as a compound of net 1 , but only with a cross-reference to the independent lemma network.

Further simplification of the structure
The target of this restructuring was compounds, especially in lemmas containing an extensive list of them.Compounds that manifested in the two ways described below were removed from their original list and made into independent lemmas.The first group identified were those compounds which had achieved 'critical mass' by having their own compounds or derived compounds, i.e. compounds closer related to the preceding compound than the original lemma, or by having a goodly collection of example phrases of their own.The second group of compounds to be moved out were those that resemble derivatives more than compounds, especially those in which the joined element cannot readily be conceived of as an independent word.The lemma nood (Pharos One, p. 376; Concise, p. 394) provides examples of both these kinds of compounds.Compounds of the first group that were made into independent lemmas are noodhulp (with an example phrase and three of its own compounds), noodlot (with several example phrases, a compound and two derivatives of its own) and noodsaak (with a verb homograph and three derivatives, each with an example phrase), while noodwendig (with an adverb homograph and a derivative) of which the etymology as a compound has become rather opaque, provides an example of the second group of compounds identified above.
Several more examples of the restructuring of compound lists can be seen at mee (Pharos One, p. 343; Concise, p. 358) and saam (Pharos One, p. 494; Concise, p. 518).However, both these sets are intricate -as examples -because of the concurrence of compounds of the variant prefixes mede-and same-.Where compounds of mee or saam were made independent, the variant derived compounds with mede-or same-were added to the entry (with cross-referencing from their alphabetically correct place).See for example the Concise entries of meedeel and meeding, and of saamdrom, saamflans, saamhang, saamkom, saamleef, saamloop, saamroep, saamsmelt, saamsnoer, saamstel, saamsweer, saamtrek, saamval, saamvat, saamvloei, saamvoeg and saamwerk.Extensive reference to the electronic archives was required to determine by means of frequency counts which of, for example, mededinging or meedinging, or sametrekking or saamtrekking, should receive precedence.

Conclusion
When Pharos Dictionaries was established in 1996, its first priority was to develop a replacement for the TW, a reference work which had set the standard for bilingual AEEA dictionaries and was widely used, but was in danger of becoming badly outdated.In 2005 Pharos released the much anticipated Pharos One, whose status as a worthy successor to the TW was soon recognised.In 2006 it was honoured by both the Afrikaanse Taal-en Kultuurvereniging (with a Woordwys Prize) and the South African Translators' Institute (for Outstanding Translation).
The outstanding features of Pharos One are its user-friendliness and accessibility, its wealth of modern material and the sensitive way in which older material is treated.All its entries offer information about morphology, pronunciation, inflection and usage and the structure of the lemmas is open and logical.The new material places Pharos One among the most up-to-date South African reference works and enables a confident interchange between modern English and Afrikaans.The collection of older material in Pharos One is comprehensive and extensively labelled, particularly regarding words and expressions which may have become offensive.
Despite its enhanced features, Pharos One has yet to equal the overall sales rate of TW8.This could be due to the size and price of Pharos One, but the significantly reduced role of Afrikaans in the present-day South Africa must have a decisive impact too.To counter the first-mentioned factors Pharos Dictionaries decided to produce a scaled-down version of Pharos One.This one-third reduction became the Concise, achieved by culling from Pharos One dated material, overly technical vocabulary and words and expressions which are particularly British, American or Australian.With dimensions of 18 cm x 25 cm x 5 cm and a mass of 1,8 kg, the Concise is considerably smaller and more affordable than Pharos One.
The average monthly sales of the Concise, released in 2007, are steady and match those of Pharos One.Significantly, this means that the combined average monthly sales of Pharos One and the Concise are now on par with those of TW8 over its entire 20-year print run.In this respect the production of the Concise has been justified: not only does it offer the user the best of Pharos One at a reduced price, its presence in the market has augmented the sales rate of Pharos's comprehensive AEEA dictionaries to a level comparable to that of TW8.This is no mean feat, considering that the TW benefited hugely during the time in which Afrikaans was one of only two official languages.
These developments demonstrate not only that the work of Pharos's editors is held in high esteem by their peers in cultural and language bodies, but also that the general dictionary-buying public continues to recognise the enduring quality of dictionaries bearing the Pharos brand.
Note how the example phrase and its translations in the TW8 excerpt have been expanded in Pharos One to include both subject and object.This is largely the legacy of the work done by Joubert in his Bilingual Phrase Dictionary, from which Pharos One borrowed extensively.
These are the expressions with corporal punishment as theme.Examples from Pharos One: iem.op sy baadjie gee give s.o. a hiding, dust/trim/warm s.o.'s jacket; iem.op sy bas gee tan/whip s.o.'s hide, give s.o. a (good) hiding/licking; 'n afgedankste/deftige/gedugte pak (slae) a sound beating; a good/sound hiding; iem.'n groot pak gee beat s.o.hollow; iem.'n helse pak (slae) gee beat the hell out of s.o.; iem.('n pak/drag) slae gee give s.o. a beating; give s.o. a hiding/tanning; jy kan maar jou lyf vetsmeer there's a rod in the pickle for you, prepare (yourself) for a hiding.
The lemmas run from Nemesis to netting.They are on p. 1183 in Pharos One, from top left to bottom right, and in the Concise on p. 1231, two-thirds down the left-hand column, to just over halfway down the left-hand column on p. 1232, which is slightly less than one Concise page in total.This example demonstrates the successful reduction of a threecolumn page in Pharos One to a two-column page in the Concise.