On the Metalexicographic Genre of Dictionary Reviews, with Specific Reference to LexicoNordica and Lexikos

: Wiegand (1984) introduced dictionary criticism as a formal component of a general theory of lexicography. Since then many scholars have focused on various aspects of dictionary criticism. In this article a distinction is made between two major types of dictionary criticism, i.e. the review of metalexicographical literature and the review of dictionaries. In the review of dictionaries different types can be identified, i.e. a short discussion without a critical evaluation, a more comprehensive review, a review article, reaction to a review or a review article, a peer review and the evaluation of a dictionary as a part of a bigger metalexicographical contribution. This article gives a brief discussion of the scope of dictionary criticism and looks at some criteria for dictionary reviews before focusing on aspects of dictionary reviews in the Scandinavian journal of lexicography LexicoNordica and the South African lexicography journal Lexikos . A comparison is made between the reviews in these two journals, identifying striking similarities and differences. It is also shown that the reviews in Lexikos frequently refer to topics that are not as relevant to LexicoNordica . This is due to the fact that dictionaries reflect something of the linguistic and cultural environment where they are published. Proposals are made to enhance the quality of reviews in Lexikos and to help ensure a stronger theoretical approach to reviews as a genre that elevates the metalexicographic discourse and improves the dictionary culture of a given community.


Introduction
In the development of lexicography as an academic discipline different theoretical approaches have come to the fore, focusing on a variety of features and components relevant to lexicographic theory.One of the significant early contributors in this regard has been the German metalexicographer Herbert Ernst Wiegand. Wiegand (1984: 15) already introduced the different subsections of metalexicography, as being (1) the history of lexicography, (2) a general theory of lexicography, (3) research on dictionary use and (4) dictionary criticism.Wiegand (1989: 262) uses the term Wörterbuchforschung (dictionary research) and allocates the following four research areas as subfields of the broader research field: (1) research into dictionary use, (2) critical dictionary research, (3) historical dictionary research and (4) systematic dictionary research.In his further research Wiegand has dealt with some aspects of the concept of critical dictionary research, cf.Wiegand (1993), but not nearly in as much detail as he has devoted to some of the other subfields.Wiegand (1998) indicated that critical dictionary research would be a topic in the envisaged second volume of his comprehensive book Wörterbuchforschung but the second volume has not been published.In the first volume (Wiegand 1998: 264) he does make the apt remark that scientific dictionary criticism is the mental/intellectual weapon in conflicts between commercial lexicography and dictionary research.
In the metalexicographic literature discussions that have been devoted to dictionary criticism have often been fairly brief but there are also more in depth discussions of dictionary criticism that can be found in, among others, Ripfel (1989), Nakamoto (1994), Bogaards (1996), Hartmann (1996), Chan and Taylor (2001), Bergenholtz (2003), Nielsen (2003), Swanepoel (2008), Engelberg and Lemnitzer (2009), Svensén (2009) and Schierholz (2015).Engelberg and Lem-nitzer (2009) offer a thorough discussion in which they focus on criticism of both printed and e-dictionaries.In spite of all the publications dealing with dictionary criticism, the topic has still not been dealt with as comprehensively as one would have wanted it to be done.Even a publication like the three volume international encyclopedia of lexicography (Hausmann et al. 1989(Hausmann et al. -1991) ) offers scant attention to dictionary criticism, albeit that one contributor, i.e.Osselton (1989), does focus on dictionary criticism in his chapter.Also the supplementary fourth volume of this publication (Gouws et al. 2013) includes but a single chapter that explicitly deals with dictionary criticism, i.e.Swanepoel (2013).
Although Wiegand (1984;1989) does not analyse this concept and the scope of the term dictionary criticism in detail some of the above-mentioned contributions, e.g.Engelberg and Lemnitzer (2009) and Swanepoel (2008) identify and discuss various subsections of this concept.
Renewed recent interest in dictionary criticism resulted in two conferences in April and June 2016 with dictionary criticism as theme, hosted respectively by the University of Silesia (Katowice, Poland) and the Friedrich Alexander University of Erlangen-Nürnberg (Erlangen, Germany) under the auspices of the consortium of EMLex, i.e. the European Masters in Lexicography, a prominent new role player on the international lexicographic scene.Proceedings of these conferences will be published towards the end of 2016 in the book series Lexicographica Series Maior.
This article gives a brief discussion of the scope of dictionary criticism and looks at some criteria for dictionary reviews before focusing on aspects of dictionary reviews in the Scandinavian journal of lexicography LexicoNordica and the South African lexicography journal Lexikos.A comparison is made between the reviews in these two journals, identifying striking similarities and differences.

The scope of dictionary criticism
When dealing with dictionary criticism as one of the components of the broader fields of dictionary research and metalexicography it is important to have a clear understanding of the scope of this research field.Looking at the metalexicographic discussion it is clear that the scope of the concept of dictionary criticism does not always go beyond the mere review of dictionaries.The current article assumes that dictionary criticism refers to a much wider lexicographic activity that includes reviews of both metalexicographic literature and dictionaries.In both these categories a distinction is made between (1) a short description without a critical evaluation, (2) a review directed at a publication as a whole, (3) a review article that offers a more comprehensive and scientifically in depth discussion, (4) response by the author of the metalexicographic work or the dictionary to a review or a review article, (5) peer reviews, e.g. of metalexicographic publications and conference abstracts, and (6) the evaluation of a dictionary or aspects of one or more dictionaries as part of a bigger metalexicographical publication.
Category ( 1) is often only a brief announcement of a new dictionary, often found in a non-scientific publication like a newspaper but also in scientific journals as a lexicographic news snippet, with a very restricted discussion of the new product, whereas category (2) gives a more comprehensive discussion, either in a scientific or a non-scientific publication, where the reviewer has the opportunity to give a well-motivated assessment of the reviewed dictionary or metalexicographic publication.In dictionary criticisms belonging to category (3) the review article offers an in depth discussion of either the dictionary or metalexicographic publication as a whole or one or more topics selected from the relevant publication for a comprehensive and critical discussion.An example of such a criticism is Swanepoel (2014) with its focus on the distinction between criteria for the description and evaluation of the design features of dictionaries, and the content, structure and operationalization of lexicographic evaluation criteria with regard to the Woordeboek van die Afrikaanse Taal, in his review article of volume 14 of this dictionary.
Contributions in category (4) are not found that often but this category represents a component of dictionary criticism that should be expanded because it leads to a critical and bidirectional discussion of lexicographic works.Authors of reviewed publications should get the opportunity to respond to reviews of their work -acknowledging the advice they receive from reviewers and motivating some of their decisions.A good example of such a response is Gove (1962) which gives a response of the editor of the Webster's Third New International Dictionary of the English Language (Gove 1961), presumably the most severely criticised dictionary, cf.Sledd and Ebbitt (1962), to the remarks of some of his critics.A more recent example of this type of dictionary criticism is the response by Botha (2005), editor-in-chief of the Woordeboek van die Afrikaanse Taal, to the critical review article by De Schryver (2005) and the subsequent response to the response (De Schryver 2005a).
Before publishing reviews of lexicographic publications journal editors will do well to submit these reviews to the authors of the reviewed publications to solicit their response and publish it along with the review.This could ensure a more balanced presentation of dictionary criticism with regard to a specific publication, but it will also stimulate the metalexicographic discourse.
Peer reviews, category (5), is a type of lexicographic criticism that has not received the necessary recognition as fully-fledged scientific contributions, cf.Bergenholz and Gouws (2015).One of the issues that could be re-assessed in the writing of this type of criticism is the tradition of anonymous peer reviews.Where a reviewer knows who the author of a publication is (either an abstract or a paper) and the author of the publication knows who the reviewer has been the nature and extent of comments and responses could perhaps take cognizance of the relevant cotext and this could perhaps eventually lead to a significantly enhanced publication.As it is the case with category (5) the importance of the contribution of critical discussions of dictionaries belonging to category (6) has also been neglected in metalexicographic literature.Many text books on lexicography or metalexicographic papers contain critical remarks and often more comprehensive discussions of certain aspects of specific dictionaries or the treatment and presentation of specific data types in certain dictionaries.Albeit that such a discussion might only focus on the way in which a specific dictionary treats one data type in a single dictionary article, it has to be regarded as a significant contribution to the field of dictionary criticism.Metalexicographers giving this kind of criticism in their publications need to be aware of and should also adhere to the criteria applicable to other forms of dictionary criticism.
This article gives a restricted discussion of reviews, looking only at reviews of dictionaries, with a focus on only dictionary criticism belonging to category (2) and category (3).No further reference will be made to the criticism of metalexicographic literature; this calls for a separate article.

3.
Establishing a dictionary criticism culture Hausmann (1989: 13) introduced the concept of a dictionary culture, i.e. society adapting to lexicography e.g. by becoming more familiar with dictionary types and dictionary using skills.Gouws (2013) argues in favour of a comprehensive dictionary culture that will not only focus on lexicography dealing with languages for general purposes but also lexicography of languages for special purposes.According to Gouws (2016) such an expanded and refined version of a dictionary culture could be further adapted to include a positive awareness of the need for and importance of a critical approach to dictionaries.A society characterised by a sophisticated and comprehensive dictionary culture will acknowledge the importance of dictionary criticism and will see this genre as a significant method of guiding users to make informed choices when consulting and buying dictionaries.Stronger theoretically-based reviews will lead to an enhanced level of dictionary criticism and therefore to a broadening of the metalexicographic discourse.The current article briefly looks at some aspects of dictionary reviews before referring to a comprehensive study of dictionary reviews in the journal LexicoNordica (Bergenholtz 2003).Some of the outcomes of this investigation are then briefly compared with those of a much more limited investigation of dictionary reviews in the journal Lexikos.

Reviews and peer reviews as form of quality assurance
Reviews always deal with new works.In the evaluation of older publications one does not have a review but rather a general discussion or historical presentation.
It is quite natural that a single scientist or even a lexicographic team will rejoice over a positive review of their work -and they will be much less excited about a negative review.But even a few points of well-founded criticism often function as a call for methodological or stylistic changes in future http://lexikos.journals.ac.za publications or revisions of the same work.This also applies to dictionariesboth with regard to the correction of concrete errors, e.g.orthographic mistakes, and underlying theoretical problems.Lexicographers should regard it as part of the responsibility of reviewers to give an honest and critical assessment of the quality of the dictionary and they therefore usually appreciate this kind of feedback that can enhance the quality of their future work.
The influence of reviews may never be underestimated and a review may even compel the publisher to withdraw the work from the market and stop any further sales.But even if a publisher does not opt for such a drastic decision after a devastating review, such a review can have a negative influence on further sales of the dictionary.The first three volumes of the six volume Brockhaus/Wahrig (Wahrig et al. 1980-84) received extremely negative reviews, cf.Wiegand andKučera (1981, 1982) and Ringguth (1982).Albeit very negative, these reviews were published in journals for linguists and due to the very restricted reader group of these journals the reviews did not have much influence.However, a discussion of these reviews in Der Spiegel, a popular German weekly magazine with more than a million copies sold weekly at that stage, most probably caused a dramatic drop in sales of the Brockhaus/Wahrig with the end result that the publishing house eventually had to withdraw the six volume dictionary from the market.
In this case it had not been the intention of the reviewers to harm the publishing house but, in order to achieve future quality assurance, to make them aware of deficiencies and mistakes in the dictionary and of problems in the practice of copying from another dictionary.However, reviews may sometimes not primarily focus on quality assurance but may rather endeavour to discredit a given publication.In another, non-lexicographic, field this kind of approach presumably prevailed in the work of four different reviewers of Bjørn Lomborg's The Sceptical Environmentalist (Lomborg 2001) that were published in the journal Scientific American.A central theme in the book by Lomborg was his scepticism, based on statistical data, regarding the forecast of a sudden occurrence of global warming, caused by humans.Lomborg specifically criticised the huge sums of money allocated to avoid this alleged warming and the fact that a lot of money would still have to be spent in future.Lomborg argues that the money could rather have been used to improve the assistance and self-assistance of the poorest people in the third world.The interpretation of some data was questioned in four extremely negative reviews.Without referring to other positive reviews in other journals these reviews were used as main argument in a complaint submitted to the Danish ethical commission for scientific dishonesty.The complaint explicitly stated in a written motivation its opposition to Lomborg's envisaged appointment as director of a new environmental institute.Lomborg was appointed but a new complaint was launched to the Commission for scientific dishonesty.In spite of protests from several hundred Danish scientists he was convicted of scientific dishonesty.Consequently Lomborg withdrew from the position that he obtained as director of the environmental institute (cf. the article for " Lomborg" in Wikipedia).
From this it is clear that reviews can have a significant influence on the reviewed publication, on future works of the same author or the same team and even on personal career possibilities.Reviews can also be employed as weapons in scientific arguments.A reviewer cannot prevent such misuse but by adhering to the methodological and ethical rules that are discussed in the following sections of this article the reviewer can at least try to produce a review that does not provide any explicit arguments to assist the misuse thereof.
Dictionary reviews can also have a positive influence.The comprehensive multi-volume Afrikaans dictionary the Woordeboek van die Afrikaanse Taal (the WAT) had been the target of extremely negative criticism, expressed in a number of reviews of especially volumes 6 and 7, cf.Grobler (1978), Combrink (1979), Odendal (1979) and Gouws (1985).As a result of this criticism the editorial staff of the WAT organised a special seminar to which metalexicographers and linguists were invited.The purpose of the seminar was to discuss the various points of criticism directed at the WAT.These discussions led to a completely revised editorial policy that resulted in much better lexicographic work in the subsequent volumes of the WAT.In his review article, focusing on volume XIV of the WAT, Swanepoel (2014: 378) argues that the compilers of the WAT have relied on input from, among others, dictionary reviewers.These reviews have enhanced the quality of the WAT (see also Gouws 2016).
In spite of this important role of reviews for individual scientists and science as a whole one unfortunately notices in the internal evaluation of research at universities an increasing lack of consideration for reviews and peer reviews.A single example is given that is typical of universities in Europe, Asia, Africa and North and South America.Table 1 shows how credits were previously obtained at the Aarhus School of Business in Denmark -also for reviews.The University had a system according to which credits were allocated to different types of publications.Each member of the academic staff had to earn a certain minimum number of credits per year and reviews contributed to achieving the prescribed aims.Today no credits are given for either reviews or peer reviews.Academics are rather requested to refrain from these activities in order to write more papers in peer-reviewed journals.Even when one includes reviews and peer reviews in your research outputs they do not count anything in the evaluation of scientific activities.This tendency at a Danish university is symptomatic of a worldwide tendency that is extremely disadvantageous to science.Fortunately, in spite of this situation there still are researchers who are willing to write reviews and peer reviews and who try to do it as good as they can.
A change in the way reviews and peer reviews are counted is much needed.Reviews are important products of research, also of lexicographical research, for the reviewed scholar anyway, but reviews and peer reviews should count more in strengthening the reputation of lexicographers, also by being recognized as fully-fledged research outputs in the research evaluation process of universities.
It is not only important to recognize the role of reviews in the process of quality assurance; it is also important to acknowledge their role in lexicographic curricula, in the formulation of lexicographic theory and as a topic in metalexicographic discussions.The curricula of academic programmes and courses in lexicography should give more attention to dictionary reviews.Criteria for dictionary reviews need to be developed and embedded in any lexicography course and potential dictionary reviewers need to be made aware of the relevant criteria.

5.
What characterizes a good review?
The following proposals for reviewing dictionaries can be expanded with many more.We only give those that we regard as essential for the review of dictionaries.
Proposal 1: The review of a dictionary has to be fair In principle a publishing house benefits from every discussion of any of their dictionaries.Obviously publishing houses hope for positive evaluations but even a negative reference to a dictionary increases its exposure and makes it better known.Although lexicographers are pleased with positive reviews, negative comments can have the advantage of helping them to improve the next edition of the dictionary.A question that does arise, is whether all dictionaries need to be reviewed or only, for example, scientific dictionaries, cf.Wiegand (1998: 40), where the review is published in a scientific journal.Should one completely refrain from reviewing non-scientific or even bad scientific dictionaries?Both these questions can clearly and unambiguously be answered in the negative.Each and every dictionary, good or bad, is an instrument that real users can use in real situations.The user has the right to receive assistance and guidance with regard to the purchasing and use of a dictionary.Metalexicographic discussion in general can also benefit from discussions of mistakes made in the lexicographic practice.Lexicography deals with dictionaries; not only good dictionaries.Therefore all dictionaries need to be candidates for reviews.The type of dictionary and the potential target user of the dictionary, the source where the review is to be published and the envisaged reader of the review will necessarily have an influence on the nature and extent of the review.Reviewers should be fair by not applying the same approach in their reviews of dictionaries from different types with different target user groups and different genuine purposes, cf.Wiegand (1998: 298).
No matter where it is published, what dictionary is reviewed or who the target readers will be, any review has to be fair.This is most likely to happen when the reviewer employs criteria for the methodology of reviews, cf. for example Tiisala (2000, 271f).Tiislala discusses dictionary reviews where the reviewers had been impressionistic.He acknowledges that this can happen very easily when reviewers only rely on random sampling of dictionary material for their criticism.A set of criteria that can be applied systematically can lead to criticism that can improve the quality of further editions of a dictionary.Each reviewer can formulate such criteria that can lead to a fair assessment of a dictionary but one can also use existing proposals, for example Steiner (1984), Leisi (1993) or Bergenholtz (2003), whose proposals are given here in a slightly altered version: -A review should give a brief account of the contents and the extent of the dictionary.But it has to be more than an account.It should also give a real impression of and create interest in the dictionary.-A review should not be too long but should have room for a presentation of some concrete dictionary articles.-A review has to include an analysis and an evaluation of the dictionary.This should be motivated by the reviewer by means of a clear indication of the grounds for the evaluation.-The review should be both an evaluation and a user orientation.
-Reviewers have to identify themselves with the editor without refraining from criticism.-The criticism should be positive by presenting clear alternatives for the criticized dictionary articles.
Being fair should also imply that a review is factually correct.It has to cite and summarise correctly without omitting important issues and without interpolations for which the reviewed work does not offer any concrete data.A review needs to be both critical and fair.In addition, and this is important, the review also has to be subjective to a certain extent, i.e. it has to present a personal evaluation.A review that does not contain a personal assessment by the reviewer on aspects of the work he/she is reviewing is not a real review but at the most a documenting description.A review should be balanced and should neither be completely objective nor completely subjective.A review is a subjective text, http://lexikos.journals.ac.za based on objective criteria, cf.Kassebeer (2015).This is how it can be explained that good reviews may present opposing assessments of the same work, and therefore it is important to look at different reviews when utilizing them for quality assessment.When evaluating reviews it is also important to know who the reviewer is and who the target readers of the review are -this may necessarily influence the nature of the review.These proposals could be formulated more briefly, as for example in Ripfel (1989: 31) or Bergenholtz and Frandsen (1997: 48): A review should (1) describe, (2) evaluate, (3) motivate the evaluation and (4) recommend or discourage the use of the dictionary.Proposal 2: The editor of a journal or a book with dictionary reviews should ensure that the review is fair Editors can follow this proposal if, from the outset, they do not invite potential reviewers where they know that there are close positive or negative relations between the lexicographer and the reviewers.In addition, an editor should not accept reviews for publication when they are in clear breach of the given methodological and ethical demands.Finally, the editor should guarantee the possibility that the lexicographer of the reviewed dictionary has the opportunity to respond to both good and bad reviews with a contribution that could be as comprehensive as the review itself.This should preferably be in the same volume of the journal in which the review is published.If this is not possible, then it should be published in the subsequent volume.

Proposal 3: A reviewer should employ ethical considerations relevant to reviews
A review can only be fair if it adheres to certain ethical demands.Wiegand (1993) uses a biblical metaphor to present a series of these "ethical commandments."We paraphrase them here with regard to dictionary reviews: -One has to demand ethical claims from dictionary reviews.
-A dictionary is reviewed and not the person that compiled the dictionary.A dictionary is not a person but the lexicographer is a person.Reference to the person of the lexicographer or lexicographers and their scientific qualifications has to be avoided -unless it is a reference to generally known facts that readers can also find elsewhere.-One should not annihilate the dictionary and thereby its lexicographer.
The commandment: 'You shall not kill!' also applies to reviews.-One should not give too much praise when it is only done to win friends or to support your career.-Although reviewers should separate the wheat from the chaff they should not only look for the wheat and ignore the chaff.One has to look for the wheat and when the reviewers do not find it they should contemplate whether they are blind in one or in both eyes.
-It is not that much of an achievement to identify a missing comma or a single typing error.-A review should not be a platform for self-promotion of your own dictionary or scientific contributions.Praise in your own mouth stinks.-A mocking reviewer is like a professor that ridicules his colleagues in his lectures.He may earn the applause from the audience but also the wrath of the gods because he ridicules a lexicographer that cannot defend himself.

Proposal 4: Each dictionary user can review a dictionary
Many potential reviewers are asked by the editor of a journal to write a review.
In exceptional cases people also submit unsolicited reviews.These two situations confront the lexicographer with the question as to who is the best suited to write the review of a given dictionary.Does one have to be an excellent metalexicographer with substantial knowledge of the language(s) of the dictionary or an experienced practical lexicographer or trained linguist?Many people are of this opinion and would therefore criticise a review when realising that the reviewer is not a well-known expert of the language(s) or an expert in the field of metalexicography/lexicography or linguistics.This criticism is often expressed by the lexicographer of the reviewed dictionary.It also happens that reviewers apologetically remark in the introduction to reviews that they have written the review without being a language expert.The criticism mentioned here is unfair if the reviewer has not claimed in the review to be such an expert.The apology is unnecessary but interesting and relevant for the reader of the review but it does not diminish the quality and relevance of the review.
In reality any reviewer, whether metalexicographer, linguist or lay person, should pay special attention to the potential functions and user groups mentioned in the front matter text of the dictionary.To a certain degree a lay person can do this exceptionally well if she/he belongs to the envisaged target user group and writes the review based on her/his own experiences of using the specific dictionary.If this can be done in collaboration with a lexicographer it could be almost ideal for a thorough and relevant review.An example of such an endeavour is Pedersen and Pedersen (1996), a collaboration between a high school student and a lexicographer, or Gundersen (2002), that introduced a child as consultant for the review of a children's dictionary.
Working with the assumption that any dictionary user can be a potential reviewer, it is important that the reader of a review should not only know who the reviewer is but also what the position of the reviewer is with regard to the specific dictionary, e.g. a member of the envisaged user group, a fellow lexicographer or a metalexicographer.Reviews should reflect the position of their authors and the assessment of reviews should be done with regard to the specific status and role of the reviewer in terms of the reviewed dictionary.http://lexikos.journals.ac.za

6.
Reviews in LexicoNordica Bergenholtz (2003) presents a comprehensive analysis of reviews of dictionaries published in the Scandinavian lexicographic journal LexicoNordica for the period 1994-2002.Some of the results of his research and some of the tables he used will be given in the following paragraphs to present a basis for comparison when looking at results from the South African lexicographic journal Lexikos.Coming from reviews in two lexicographic journals the results from LexicoNordica and Lexikos are ideal for a comparative study.The criteria applied in Bergenholtz (2003) have therefore also been used in the analysis of Lexikos.
The following table from Bergenholtz (2003) gives the number of reviews published in LexicoNordica in each year of the period investigated: These reviews, a total of 110 over a period of 9 years at an average of 12,2 per annum, were directed at different types of dictionaries, e.g.monolingual general language dictionaries, bi-or polylingual general language dictionaries, bior polylingual specialised dictionaries and monolingual specialised dictionaries, lexicons and encyclopaedia.

LexicoNordica
The core of the analysis by Bergenholtz which will also be compared with Lexikos regards the topics referred to in these reviews.The following table illustrates these categories and the number of references to them in the reviews in LexicoNordica.The total number of references exceeds the number of reviews because a single review usually contains a discussion of more than one aspect of the specific dictionary and often more than one reference to any given category in the list below.As an example: different references can be found in different sections of a single review to the same aspect, e.g.grammar.Each one of these references has been counted individually.Consequently in e.g.ten reviews the number of references to grammar exceeds the number of reviews under discussion:

Types of reviews
In section 2 of this article reference was made to different types of reviews, including (1) a short description without a critical evaluation, (2) a review directed at a publication as a whole, (3) a review article that offers a more comprehensive and scientifically in depth discussion, (4) response by the author of the metalexicographic work or the dictionary to a review or a review article.
The editors of Lexikos, as reflected in the presentation of different categories of contributions in the table of contents, make provision for two types of dictionary reviews.In the table of contents they are classified as Reviews and Review articles, with the latter infrequently occurring in the various volumes of Lexikos.
The Lexikos category of Reviews includes the categories (1) and ( 2) mentioned above -unfortunately without any distinction that could inform the reader of the nature, scope and extent of the discussion contained in the review.Just as is the case with dictionaries, dictionary reviews should also have clearly identified envisaged target readers.This target readership is primarily determined by the journal or other publication in which a given review is included.Where reviews in a newspaper are directed at a general reader group, a review in a scientific journal should be directed at readers who are semiexperts and experts in the relevant subject field of that specific journal.A dictionary of physics that is reviewed in a journal of physics should be directed at experts and semi-experts in the field of physics.When the same dictionary is reviewed in a lexicographic journal, the focus is not necessarily on the contents in terms of the needs and demands of physicists because physicists do not typically read a lexicographic journal.Such a review will be directed at the target readers of the specific journal, i.e. people interested in the field of lexicography.The scope of their interest could include a discussion of the contents of the dictionary of physics but could also make provision for remarks regarding e.g. the structures, functions, user-perspective, etc. of the given dictionary; aspects that would most probably not be of interest to the physicists.This target reader group of a review should necessarily influence the approach followed by the specific reviewer and the nature of the review.
In a scientific journal like Lexikos one would expect the majority of the reviews to be directed at members of the lexicographic community, i.e. experts and semi-experts in the field.These typical readers of Lexikos would typically expect a high level review that goes beyond a brief introduction of the given dictionary.Unfortunately this is not the case in many of the reviews.Too many of these reviews belong to the above-mentioned category (1) and do not contribute significantly to a substantial critical lexicographic discourse.Fortunately there are also ample reviews that are more comprehensive as well as review articles that play an important role in developing dictionary criticism as a component of a general theory of lexicography.By allocating a section in the jour-nal to reviews and by having a dedicated review editor on the editorial board, the editors of Lexikos ensure that this journal does not only give a fine reflection of new dictionaries but it also stimulates the metalexicographic discussion.
The quality of the contribution of Lexikos to the field of dictionary criticism could be enhanced by a more structured approach in the writing of reviews.When soliciting dictionary reviews the review editor could consider the possibility of giving the potential reviewers a brief of the kind of review needed for this journal, cf.Gouws (2016).Too often the reviews do not really live up to the expectations of the target readers of Lexikos.Such a brief that deviates from a random approach to say something about a dictionary could also help to develop the expertise of dictionary criticism.This would yet again elevate the level of the metalexicographic discourse.

Focal points in the Lexikos reviews
Looking at reviews in Lexikos, volumes 1-25, and employing the set of criteria used by Bergenholz (2003) it is interesting to note both similarities and differences when comparing these reviews to those in LexicoNordica.Compared to LexicoNordica's 110 reviews in 9 years at an average of 12,2 per annum, the 25 volumes of Lexikos contain 126 reviews at an average of only 5,04 per annum.
There is no fixed number of pages allocated to reviews and no clear indication of a trend in terms of the frequency or extent of reviews or review articles in the 25 volumes.Before reflecting on a comparison between the two journals it is interesting to look at some other topics in the reviews in Lexikos.Albeit that Lexikos is an international journal it is based in South Africa.Many dictionaries from South African publishing houses are submitted for reviews and the editors have to respond to these submissions.Dictionaries are not only compiled for specific target users but also to be used in specific linguistic and cultural environments.It is to be expected that South African dictionaries will reflect aspects of the South African multilingual and multicultural situation and these issues are often addressed in the reviews.This implies that some issues come to the fore in the reviews in Lexikos that have no or a far lesser impact in the reviews in LexicoNordica.
A number of reviews include references to a typical lexicographic topic like the treatment of homonymy and polysemy.Due to the need for a wideranging variety of dictionary types in South Africa, the position of a given dictionary within the typological spectrum is a frequent point of discussion in the reviews.Within some of the African languages lexicographers need to make definite choices with regard to the lemmatisation procedures to be followed in their dictionaries.In this regard the distinction between a word or a stem lemmatisation approach is relevant, and this is a topic that receives coverage in many reviews.The occurrence of dialectal forms and the reality of regional differences in the South African languages are reflected in reviews.Within a multilingual environment language contact inevitably leads to language influence, and therefore the occurrence of loan words and other borrowings should be reflected in dictionaries.The reviewers are keen to identify loan words that have been included as lemmata in dictionaries.The influence of the political situation on the languages of South Africa and the way in which it is reflected in dictionaries also come to the fore in reviews.Sensitivity for language use that can be offensive is often expressed.This applies to words from the political arena but also language use that could be seen as racist or sexist.Reviewers show their readers how the reviewed dictionaries live up to the expectations of dictionary users to find an objective reflection of the specific language treated in the dictionary.
The following table gives a comparison of the percentages of references to the different categories in the two journals: Each pairing of categories in this table could be discussed in detail.This will not be done in this article.Only a few remarks will be made with regard to some striking similarities and differences.Many reviewers are keen to say something about the number of words or new words or the types of words and expressions included in a dictionary.Therefore it can be expected that the lemma selection should be a frequent topic of discussion -the topic most frequently referred to in the reviews in both journals.Aspects about semantic and encyclopaedic knowledge also have a comparable high standing in these tables.It is often, and rightfully, said that few users read the front matter texts, including the users' guidelines text in a dictionary.Reviewers in both journals pay significant attention to the outer texts, i.e. the front and back matter texts.This may be seen as an attempt on the side of the reviewers to help to establish a dictionary culture where users are aware of the fact that a dictionary contains more than just the central list, i.e. the alphabetical section.

Lexico
This table also shows striking differences, e.g. in categories 5, 8, 33 and 34.These differences can be explained by taking cognizance of editorial traditions in the two journals and by prevailing approaches to dictionary reviews on the side of the reviewing community.
Category 5, the price of the dictionary, shows the biggest deviation between the two journals with LexicoNordica scoring 0,9 and Lexikos no less than 11,1 -the second most frequent category in Lexikos.This is due to a template introduced in dictionary reviews where the heading of the review includes information on the title, author, publisher, etc. and also gives the price of the dictionary.In category 8 the reviews in LexicoNordica refer much more often to other dictionaries, compared to the reviews in Lexikos.This could be seen as typical of a more developed culture of dictionary criticism where the reviewed dictionary is not seen in isolation and readers have the opportunity to assess the dictionary by comparing some of its features to those of other dictionaries.
Categories 33 and 34 show that reviews in both journals give an assessment of the dictionaries but such an assessment more frequently occurs in the Lexikos reviews.According to proposal 1 (The review of a dictionary has to be fair) given above, a review should give an evaluation of the dictionary.This seems to happen more often in Lexikos than in LexicoNordica.However, the positive assessment (7,7%) in Lexikos does not seem to realistically reflect the quality of dictionaries.Too often too many reviews refrain from a negative overall assessment even when individual categories are criticised negatively.Reviewers should work with the idea that criticism can and should be both positive and negative and that it is the responsibility of the lexicographer to give an honest assessment of the dictionary.
The largely comparable percentages in this table indicate that the categories selected could be regarded as representative of what one typically finds in a dictionary review published in a scientific journal of lexicography.It would be interesting to see the results of a similar comparative study of other major journals of lexicography.

Conclusion
As an established international journal Lexikos will do well to continue with its policy of having a review editor and publishing both reviews and review articles, reviewing both dictionaries and metalexicographic publications.The importance of reviews could be emphasised by allocating more pages to this component of the journal's contents.The relevance of reviews as part of the academic discourse could be elevated if lexicographers of the reviewed dictionaries or authors of the reviewed metalexicographic works are invited to respond to the reviews.The quality of reviews could be enhanced if reviewers are given a brief with an indication of the genuine purpose of the review, cf.Gouws (2016), and the expectations of the intended target readers.These reviews should move away from a mere announcement of the dictionary to a comprehensive discussion of its lexicographic nature and contribution.When soliciting reviews the review editor could consider to invite reviews from reviewer teams, consisting e.g. of a (meta)lexicographer and a typical target user of the reviewed dictionary.These dictionary users could give a valuable input and their ideas could help the other reviewer but also the lexicographers and the lexicographic community at large to embark in a much more effective way on the planning and production of user-directed dictionaries. http://lexikos.journals.ac.za

Table 2 :
Number of reviews in LexicoNordica

Table 4 :
The following table indicates the different categories and the number of relevant references (with different references to the same category in a single review counted individually) in Lexikos: Topics referred to in reviews in Lexikos 1991-2015.

Table 5 :
Comparison of percentages in LexicoNordica and Lexikos.