In what sense is the Louw / Nida dictionary authoritative ?

In w hat sense is th e L ouw /N ida d ictionary au th o ri­ tative? T his essay deals w ith fou r fac to rs w hich m ake the Louw /N ida dictionary authoritative. These include the presentation of lexical meanings in sem antic domains, the definition of meanings, the distinction that is made betw een meaning and reference, and the layout of the d ictionary . All fac to rs a re discussed critically and examples of where the dictionary fails are given. The Greek-English lexicon o f the New Testament based on semantic domains, edited by J P Louw & E A Nida, and hereafter referred to as L& N, has received a re­ markably positive reception by the scholarly world until now. According to Reese (1988:150): ‘[T]he publication o f this attractively presented lexicon will force the scholarly guild to pay a tten tio n to the linguistic m ethodology underlying this rev o lu tionary ach iev em en t’. In a sim ilarly positive review , Silva (1989:165) maintains: ‘[T]his work has to be regarded as a prodigious step forward in the field of lexicography generally, and in the study of N T vocabulary specifically’ (see also Boers 1989; Botha 1989; Elliott 1988; Lategan 1988; Snyman 1988). I am also of the opinion that it is a major achievement in lexicographical studies of the vocabulary of the G reek New T estam en t. T he d ictionary offers a new app roach to lexical sem antics in th is field , and it is in m any ways to tally d iffe ren t from existing d ic tio n a rie s on the New T estam en t. T his gives rise to the question of how authoritative this lexicon is. • This essay was prepared for a symposium on the Louw/Nida wordbook at the annual meeting of the Society of Biblical Literature in New Orleans, November 1990. 26 ISSN 0259 9422 = HTS 47/1 (1991)

In what sense is the Louw/Nida dictionary authoritative?
W illem S V orster

A bstract
In w h at sense is th e L o u w /N id a d ictio n a ry a u th o ri tative?
T h is essay d e a ls w ith fo u r fa c to rs w hich m ak e th e L ouw /N ida dictionary authoritative.These include the p resen tatio n of lexical m eanings in sem antic domains, the definition of meanings, the distinction th at is m ade betw een m eaning and reference, and the layout of the d ic tio n a ry .A ll fa c to rs a re d iscu ssed critic ally an d examples of where the dictionary fails are given.
T he Greek-English lexicon o f the New Testament based on sem antic domains, edited by J P Louw & E A N ida, and h e re a fte r re fe rred to as L & N , has received a re m arkably positive reception by the scholarly world until now.A ccording to R eese (1988:150): '[T]he p u blication o f this attractively presen ted lexicon will force the sch o larly guild to pay a tte n tio n to th e lin g u istic m eth o d o lo g y u n d erly in g this re v o lu tio n a ry a c h ie v e m e n t'.In a sim ilarly p o sitiv e review , Silva (1989:165) m aintains: ' [T]his work has to be regarded as a prodigious step forward in the field o f lexicography generally, and in the study o f N T vocabulary specifically' (see also Boers 1989;B otha 1989;E lliott 1988;Lategan 1988;Snyman 1988).I am also o f the opinion that it is a m ajor achievem ent in lexicographical studies of the vocabulary of th e G re e k N ew T e sta m e n t.T h e d ictio n a ry o ffers a new a p p ro a c h to lexical sem an tics in th is field , an d it is in m any ways to ta lly d iffe re n t fro m existing d ic tio n a rie s on th e N ew T e sta m e n t.T h is gives rise to th e q u e stio n o f how authoritative this lexicon is.T he dictionary gives a com prehensive treatm en t of approxim ately 25 000 meanings o f som e 5 000 w ords w hich a re u sed in th e N ew T e sta m e n t.I shall lim it my discussion of the question of the authority of L & N to lexical sem antics only.
A ccording to Louw (1979:109) (Louw 1979:109).H ow successful is L & N in this H T S 47/1 (1991) regard?This question will be approached from different perspectives.Let us first deal with the arrangem ent of meaning in the dictionary.T he m ain difference betw een L A N and o th e r bilingual dictionaries on the New T estam ent is that in L & N the focus is on the related meanings of different words, and not on the different meanings tor which a particular word is used in the New Testam ent.T hat is why the many meanings for which G reek words are used in the New T estam ent are organised into ninety-three sem antic areas or domains.This is p erh ap s the m ost innovative aspect of th e dictionary since it is the first tim e in history th a t an attem p t has been m ade to arran g e the d ifferent m eanings o f the com plete New T estam ent vocabulary into sem antic domains.Friedrich (1973).T he assum ption is that L A N 's classification reflects the sem antic structure of the G reek o f the New T estam ent.Louw (1979:109) even claims that their classification is based on an 'em ic' approach, 'th at is to say, an assignm ent of meanings as they would have been classified by native speakers of K oine G reek '.In accordance with the view o f the ancients who regarded 'fire' as one of the four basic elem ents, nOp, for example, is classified as a substance and not as a process, it is argued (see also Louw & N ida 1988:xiv).
T h ere are a few basic insights, and perhaps even problem s, which have to be dealt with before we continue our discussion of the authority of the dictionary in this It is well known th a t the term s 'em ic' and 'etic' w ere first used by an th ro p o logists in the description of behaviour (see Pike 1966).T hese term s represent two standpoints 'which lead to results which shade into one an o th er' (Pike 1966:152).
W hile the etic (coined from phonetic) standpoint studies behaviour and language from outside a system, the emic (from phonemic) standpoint studies the system from inside.T h ere is a vast difference betw een th e two ap p ro ach es, especially with regard to the G reek o f th e New T estam ent, although th ere also is no reason for a dichotomy betw een the two approaches.The one is dependent upon the other.Pike (1966:153f) gives a very useful survey of the most im portant characteristics of the two standpoints.It becomes clear that, if we apply these characteristics to the study of the language of the New T estam ent, one can hardly speak o f studying the G reek vocabulary solely from an emic point of view.W hile the emic structure o f a system , fo r ex am ple, has to be d isco v ered , th e e tic stru c tu re is c re a te d .Pike (1966:153) argues.With regard to New T estam ent G reek one will have to assume two things, if one holds the view that it is analysed emically.T he first assumption is that there is an inherent system in the language, and the second is that it is possible to discover it.Both these views are problem atic.Let us deal with the latter first.
We study the G reek o f the New Testam ent, which is p art of H ellenistic G reek of that era, mainly from an etic point o f view.D ue to the lack o f native speakers and the lack of exact inform ation on the spoken language of the writers of the New T estam ent docum ents, there is no reason for any scholar to claim th at the language of these docum ents is studied o r described from inside, th a t is from the standpoint of a native speaker of that language (see also Ossege 1975:79).The problem is that we know so little about the 'language of the New T estam ent'.It is only for the sake o f convenience th a t we speak of 'New T e stam en t G re ek '.T he docum ents w ere w ritten by different authors, in different parts of the N ear E ast and Asia, at different tim es and in d ifferen t circum stances.Som e o f th e docum ents w ere w ritten by bilingual authors whose second language was G reek, o thers by authors w ho had a good com m and of the language.In w hat sense can one then speak of the sem antic stru ctu re of the G re e k of the New T esta m e n t? T hese factors c o n trib u te to the difficulties involved in having an emic view of the 'language' of the New T estam ent.
It is m ore than difficult to discover the cultural key -that is, the knowledge of the   From the perspective of cognitive semantics, 'universals', such as objects, events, and abstracts, are not regarded as common structures or elem ents of language.They are regarded as comm on strategies to classify experience.Language is furtherm ore seen as p a rt and parcel of cultural contexts (see G eeraerts 1986:198f).'Sem antic structures are taken to be nothing other than conceptualizations shaped for symbolic purposes according to the dictates of linguistic convention' (Langacker as quoted by G eeraerts 1986:211).
T he point I wish to make is that the compilers of L & N are responsible, on both em ic and etic c o n sid eratio n s, for th e ir classificatio n of re la te d m eanings into dom ains and subdom ains.They did no t discover these categories in the sem antic structure o f the G reek New T estam ent vocabulary.They created these dom ains in the light of their analysis of the related meanings of words.Once one realises this, the dictionary can be used in an interactive fashion.If one knows that the domains of L & N are possible ways of dealing with related meanings in the New Testam ent, th e u ser can sta rt using the m aterial critically.This is exactly w hat m akes L & N a u th o rita tiv e .It is no t becau se the last w ord has been said a b o u t th e relatio n betw een related meanings of different words in the New T estam ent that this lexicon is an authority that has to be followed.It is because the dictionary enables the user to see a particular word within the context of other words which are possibly related that it is to be regarded as an authority.In the words of Louw and N ida (1988; dealing with m eaning.T h ere are, however, also disadvantages in this approach to the analysis of meaning. As we have seen above, with regard to dom ains, it ap p e ars an illusion th a t sem antic fields can be clearly m arked in acco rd an ce with an in h e re n t sem antic structure.Since sem antic dom ains are fuzzy, and since th e b oundaries o f lexical meanings are also fuzzy, one can never say with certainty that a particular feature is the diagnostic co m pon en t o f th e lexical m eaning o f a p a rtic u la r w ord (see also G eeraerts 1986:112ff).M eanings overlap and are dependent on subtle distinctions and on connotative and associative elem ents, and can therefore never be analysed in a m echanical o r m achine-like m anner.T h at, p erh ap s m ore th an anything else, explains why some of the definitions of L A N are b e tte r than others, and why some are totally inadequate.A few examples may illustrate the point.
I w ould regard the definition of the m eaning of Gumáo) as 'to burn arom atic substances as an offering to G od' (53.25), for example, of xú<tio)aai as 'the process of b u rning slowly, w ith accom panying sm oke and relativ ely little glow '( 14.64), of CTeiCT)j,ó(; as 'a su d d en and severe m ov em en t o f th e e a rth ' (14.87), and all the m eanings defined in 15 as ad equate and good definitions o f the lexical m eaning of th e p a rtic u la r w ords.O th e rs, how ever, a re unconvincing and in a d e q u a te for different reasons.The definition of Panticrcfiq (53.42) as 'one who baptizes' does n ot explain the m eaning o f the word.W hat is the difference b etw een the lexical m eaning of the idioms orrpaTUx oúpáviog (12:30) and a x p a tla tou oOpai'oO (12-45)?I find the definitions 'a large group o r throng o f angels' (12:30) and 'th e stars of heaven as symbols of various supernatural powers' (12:45) unconvincing.W hat does it m ean to be 'fre e ' as in th e case o f éX euGepía ('the state o f b eing fre e ') and éXeúGepoí;® ( 'p e rta in in g to b eing fre e ')? T he vagueness o f th ese d efin itio n s becom es a problem when one com pares them with éXeú0epo<;*' (87.84), 'pertaining to a person who is not a slave, eith er one who has never been a slave or one who was a slave formerly but is no longer'.
T he fact th a t defin itio n s o f m eaning a re om itted in som e cases also causes problem s.The meaning of éKKCUxéíi) (19:14) is defined as 'to pierce with a pointed instrum ent', while the meaning of víkjcto) is not defined.Both are then translated by 'to pierce'.In the case of the latter there is an annotation -reading: 'norm ally not as serious a w ound as is im plied by éKiceuTéa)'.T his exam ple also illustrates the problem of defining the meaning o f words from the outside.A lthough the compilers hold the view th a t th e re is no such thing as w ords having com p letely th e sam e m eaning (Louw & Nida 1988:xv), they are unable to distinguish th e subtle differen-H TS 47/1 (1991) ces betw een these words that are used by the sam e author in close connection (see, however, 19.18 and 19.19).T he entries under 53.4-7 also illustrate that it is alm ost impossible to define close synonyms in the New T estam ent with the help of compon en tial analysis.W e ju st do no t have th e com petence to distinguish the subtle differences in the meanings of these words.
T he compilers are moreover, not always consistent in their definitions of lexical m eaning.A good exam ple is the definition o f w ords referring to specific festivals (see 51.5-12).In som e cases even the date of the festival is given, while in others it is not.T h e sam e ap p lies to w ords for coins in 6.76-82.In m ost cases it is said w hether the coins w ere R om an or G reek.Only in the case of Xctitóv and crcorcrp is this not done.A re these features not diagnostic for outsiders as well in these two cases?
Some A final factor which m akes this dictionary authoritative is its layout, since it compels the u ser to consider thoroughly w h eth er a p a rticu lar m ean in g is ap p licab le.It therefore has an educational function.47/1 (1991) way in which L & N presents inform ation on the lexical m eaning of New T estam ent words forces the user to make use of the information in a creative way.
A lthough L & N gives a co m p lete survey o f all th e m eanings fo r w hich th e d ifferent w ords in the New T estam en t are used, it is not com plete with regard to each case w here a particular meaning is applicable.This simply m eans that the user has to m ake full use of the inform ation provided in the dictionary.If a particular case is not m ention ed in the referen ce index, the u ser has to consult th e o th er indices.H aving decided th a t a p articu lar w ord m entioned u n d er a G reek entry might be applicable, both the definition o f that meaning and the definition of other related meanings in its im m ediate area have to be consulted.This is the only way to be relatively sure that a chosen meaning is applicable in a particular context.
Instead of going through the list of d ifferent m eanings o f a p articu lar G reek word, as they norm ally occur in an alphabetically arranged dictionary, the user is introduced the sem antic area of related meanings of different words.This has the advantage that, in addition to the different meanings of the sam e word, the user also sees the related m eanings o f different term s in the sam e sem antic dom ain.Since m e a n in g is a lso m ostly d e fin e d an d n o t p re s e n te d in th e fo rm o f glosses or translational equivalents, the user has to decide actively, with the help of sem antic inform ation provided by the dictionary, w hether a particular choice is applicable.Columbia: H ornbeam Press.

•
This essay was prepared for a symposium on the Louw/Nida wordbook at the annual meeting of the Society of Biblical Literature in New Orleans, November 1990.ISSN 0259 9422 = H TS 47/1 (1991) W S Vorster In the introduction to L & N the compilers assert that: T he principal reason for a new type o f G reek New T estam ent lexicon is the inadequacy of most existing dictionaries, which for the most part are lim ited in indicating meanings, since they depend principally on a series o f glosses.(Louw & N ida 1988:viii) R a th e r th an asking w hether th ese inadequacies have b een overcom e in this new dictionary, I decided to address the problem o f the authority o f L & N (see B arnhart 1980).By this I m ean w hether the scholarly world can rely on this dictionary as an 'authority' for the lexical meanings of the words used in the G reek New T estam ent.'A u th o rity ' is used h ere in th e sense of 'th e pow er o r right to co n tro l, ju d g e or p ro h ib it th e actions o f o th e rs' (C ollins 1986 s v).It is a p p a re n t th a t th e re are aspects on which this dictionary cannot be regarded as authoritative since, it was not designed to give inform ation on such aspects.It is, for example, not authoritative on etymology, m orphology, pronunciation, syntax, th e lexical m eaning o f each occur rence of a particular word in the New T estam ent, and the m eaning of the w ords dis cussed in their usage outside the New T estam ent, to m ention but a few things.L & N is a sem antic dictionary dealing with the lexical meanings of the G reek w ords in the New T estam ent.It can th erefo re only be authoritative with regard to the m eaning of words in the New T estam ent.T he question of authority will therefore be addres sed in this respect.It sh o u ld b e k e p t in m in d th a t L & N w as d e v e lo p e d p rim a rily fo r u se by translators.This explains th e many annotations relating to translation included in th e dictionary.T his does not, how ever, im ply th a t th e dictio n ary can n o t b e an authority or a help for o th er users such as students, exegetes, theologians and all other persons interested in the m eaning of New T estam ent words.O n the contrary.
: 'T h e p u rpose of th e W ordbook is to provide accu rate and p ractical guidance in determ in in g satisfactory equivalences fo r the G reek N T vocabulary'.This m eans that translators and exegetes are provided with 'clear descriptions of areas of m eaning o f single words, set phrases and idioms', and th a t indications a re given o f 'how equivalences o f m eaning may b e satisfactorily expressed in o th e r languages'

T
he idea of the arrangem ent o f meaning into sem antic domains is nothing novel (see G eeraerts 1986:67-148).W hat is new is the arrangem ent o f the different m ea nings of G reek New T estam ent term s into sem antic fields.TTie question we have to ask is w hat this arrangem ent is based on, and what m akes the arrangem ent of I A N authoritative.The compilers o f the dictionary assert that the sem antic domains of L A N reflect the classification of words into sem antic areas on the ground of th ree basic kinds of se m an tic co m p o n e n ts, nam ely sh a re d o r com m on c o m p o n en ts, d istin ctiv e or d iag n o stic co m po nen ts, and su p p lem en tary c o m p o n en ts (see L ouw 1979:109).A ccording to th eir view (see Louw 1979:108f; also Louw & N ida 1988:ix) these dom ains do not follow a system o f classification b ased on, for exam ple, logicophilosophical categories as in the case of R oget o r the classification proposed by regard.First of all it is necessary to decide what the compilers of the dictionary did w hen the classification of th e d ifferen t m eanings of th e G re e k N ew T esta m e n t vocabulary was m ade.D id they discover, analyse o r find th e sem antic dom ains inherent in the vocabulary, or did they create them ?28 H TS 47/1 (1991) t v S V(mter em ic system -o f th e G ree k o f th e New T e sta m en t.W hat we actually do is to co n stru e th e c u ltu ral key by studying th e lan g uag e from o u tsid e. T h is, in my opinion, also applies both to the classification of the sem antic structure of the New T e s ta m e n t v o cab u la ry in to se m a n tic d o m a in s by L ouw an d N id a, a n d th e ir d e fin itio n of lexical m eanings o f this vocabulary.T he sem an tic dom ains w ere created rather than discovered, and so was the definition of meanings with the help of com ponential analysis.Let us elaborate.The com pilers o f L & N correctly reject the structure proposed by Friedrich and have convincingly draw n atten tio n to problem s involved in the structuring o f the lexical m eanings o f th e New T e stam en t vocabulary in existing d ictio n aries (see Louw 1985b:53ff).B ut what about their own structure?A re their dom ains built on th a t inh erent in the sem antic structure of the New T estam ent vocabulary, as they presum e? TTie dom ains o f L A N obviously reflect the theory on which the dictionary is based.In th e first place they reflect th e idea th a t words can be divided into four m ain c a te g o rie s nam ely -events, objects, abstracts a n d discourse referentials.Furtherm ore, they reflect the idea that the meaning o f words can be determ ined by an analysis of th eir com ponents, and th a t this excludes the use o f encyclopaedic know ledge, to m en tio n only thw o fu rth e r asp ects.T h e re a re m o re.T h is has resu lted in a very im pressive classification o f th e d ifferen t m eanings o f re late d words into ninety-three main domains of meaning.O ne should, however, not forget th a t th e th r e e m e n tio n e d p rin c ip le s o f c la s sific a tio n fo rm th e b asis o f th is classification.In fact, they explain w hat th e com pilers discovered and how they succeeded in creating their classification.W hether these principles comply with the way in w hich first-cen tu ry G re ek -sp eak in g C h ristian s used lexical m eanings is another m atter.Som e of the dom ains may certainly overlap with the way native speakers may have understood th e relatio n betw een related m eanings.

M
aK apioq is a lso liste d u n d e r d o m a in 25 w ith w o rd s su ch as IXapoxriq, éxxjipoaúvTi, x«pó", auyxaípo), cnji/fi8op.aikxX in subdom ain 25.K: 'H appy, G lad, Joyful'.T h ere are, however, interpreters who argue that the 'b eatitudes' should be 30 H TS *7/1 (1991) re ad as apocalyptic blessings (see G uelich 1976).T h ese m ight a rg u e th a t the m eaning of jiaKÓpio^ would not be 'pertaining to being happy, with the implication of enjoying favorable circum stances', but th a t the w ord should be placed u n d er dom ain 53, 'Religious Activities'-that is, 'pertaining to being blessed by G o d '.Such examples can be multiplied.O ne soon realises th at sem antic dom ains can easily be stru c tu re d in d iffe re n t ways d ep en d in g on w h at o n e sees as th e d istinguishing com ponent of m eaning (see also G eeraerts 1986:112ff).A good example would be éctGíw.According to L A N the word m eans 'to consume food, usually solids, but also liquids'.The last p art o f the definition is included to accom m odate the use of n o ijia íu e i noíjiTii/ koI éK toO yaSaxxoc, Tty; noi(j,i/riQ oúk eaG iei: in 1 C orinthians 9:7.T o my m ind th e use o f th e p re p o sitio n ÉK w ith th e v erb has to be ta k e n seriously.1 think that coGiu) m eans 'to consume food' and that it is used in the New T estam ent for that meaning only, also in 1 Corinthians 9:7.It is furtherm ore used as a close synonym of xpcijyo) and is not used in the sense of 'eating liquids'.If I had to structure the subdom ain, I would have placed eaGuij first and then Tpúyoj and so on, in th is o rd er.T h is is ju s t to show th a t th e stru c tu rin g o f th e d o m a in s is d ete rm in e d by o n e 's in te rp re ta tio n of th e lexical m ean in g o f w ords.B oth the structure and the definition of meanings are ascribed, obviously norm ally on good grounds, not discovered.Let us go a little further.'[W]ords are rooted in social systems; they realize meanings from social systems' (M alina 1987:358).T o classify th e m eaning of G reek New T estam ent w ords into se m an tic d o m ain s p re su p p o se s th e ab ility to re c o n stru c t o rig in al c o n tex ts of c o m m u n icatio n in the first-ce n tu ry M e d ite rra n e a n w o rld .T h is is an alm o st impossible task.T he most we can do is to construct possible cross-cultural contexts of understanding.T he idea that there is an inherent sem antic structure in language, which has to be discovered and describ ed , is equally p ro b lem atic.It is b ased on stru ctu ral sem antics, which was in vogue from 1930-1975.O ne of the m ain problem s is th at too little distinction is m ade betw een knowledge of language and knowledge of the w orld.S tru c tu ra l sem an tics p re su p p o se s th a t la n g u ag e has its own sem an tic structure -which has nothing to do with encyclopaedic knowledge.In the w ords of Louw (1985a:80): 'It is ... im portant that we should add strictly sem antic dictionaries to o u r list o f d iffe re n t types o f d ic tio n a rie s'.T h is view has b e e n re je c te d as illusionary since it is argued th at the difference betw een encyclopaedic and 'p u re' sem antic knowledge does not exist (see G eeraerts 1986:187).
x): 'T he prim ary value o f a lexicon based upon sem antic dom ains is th at it forces the reader to recognize some of the subtle distinctions which exist betw een lexical items whose meanings are closely related and which in certain contexts overlap'.A second factor which m akes L & N authoritative is the definition of the m eaning of individual words.In this respect, the dictionary is totally different from all existing bilingual dictionaries on the vocabulary of the New Testam ent.T he d escrip tio n o f m eaning by way o f d e fin itio n in stea d o f in glosses and tran slatio n al equivalents m akes this dictionary unique.T h e definitions a re the result o f the analysis of the com ponents of the lexical m eaning of each individual w ord.In m ost cases th e d efin itio n s refle c t th e d iag n o stic co m p o n en ts o f the particular m eaning.In this m anner the different meanings are identified and made m o re p re c ise .T h e a d v a n ta g e o f d e fin itio n s o f m e a n in g o v e r tra n s la tio n a l e q u iv a le n ts is o b v io u s.W h ile tr a n s la tio n a l e q u iv a le n ts a re a p p ro x im a te presentations of th e lexical m eaning of a w ord in the source language, definitions are supposed to be m ore precise.i C o m p o n en tial analysis is n o t an unknow n m eth o d o f analysing th e lexical meanings of words.The advantage of defining meaning in this way is th at the focus is on those com ponents th at distinguish the m eaning of particular words from one another.It also helps to see which com ponents are shared by w ords with related m eanings and w hich are supplem entary.Ideally, this is one o f the best ways o f figurative extension, the m eaning 'd ev il' w ould also be a fig u rativ e extension.R eferences becom e meanings of words through convention.A nother, perhaps even clearer, case would be ó(j)6ciX)jó(; nourpó<;a-t>-Both meanings are regarded as idioms in L& N.They have been defined as 'a feeling of jealousy and resentm ent because of w hat som eone else has o r does' (88.165), and 'to be stingy'.Nowhere does one, however, find a reference to the belief in an Evil Eye which was comm on at the tim e of the New T estam ent in the M editerranean world, as J H Elliott (1988) has convin cingly shown.T he inclusion of oúpauó<;<= (12.16) under 'S upernatural Beings' is furtherm ore understandable, but nevertheless problem atic.Instead o f giving a definition o f the meaning of the word, an annotation explains the term .It reads 'a reference to G od based on the Jew ish tendency to avoid using the nam e or direct term for G o d '.If oúpauóq is used, like D"*ny>, as a replacem ent for the nam e of G od, one would first of all expect the term under 'Names o f Persons and Places'(93).S ince th e r e a re m any a n n o ta tio n s in c lu d e d in L & N fo r th e sa k e o f th e translator, it would have been of help had there been m ore rem arks about reference in cases w here it is possible to determ ine what words refer to.

U
sers o f lexica often tend to think th a t th e p u rp o se of bilingual lexica is to provide th e user with the m eaning o f a w ord in a particular context.It is often not realised th a t a lexicon is only an aid for the user to d eterm in e th e m eaning o f a p a rtic u la r w ord in a p a rtic u la r context.It is th e u ser, n o t th e lexicon, o r th e com pilers o f the lexicon, who has to determ ine the m eaning o f a word in use.The

I
have discussed four factors which m ake L & N an authority on lexical meanings of the New T estam ent vocabulary.T here are others.I am thinking of the im portance of figurative extensions of m eaning, the treatm en t of m eanings of word groups and especially of idioms.The fact that idioms are listed and treated m akes L & N unique.Space does not allow m e to go into these factors further.Enough has been said to in d ic a te th a t th e r e a re g o o d re a s o n s fo r th is d ic tio n a ry to b e re g a rd e d as au thoritative in d ifferen t aspects and on differen t grounds.T he p ublication is a m ilestone in the history o f New T estam ent lexicography.W orks cited B arn h art, C L 1980.W hat m akes a dictionary a u th o ritativ e, in Z gusta, L (ed), Theory and methord in lexicography: Western and non-W estem perspectives, 33-42.
Subdom ain 88, of which these w ords form p art, concerns 'M oral and E thical Q ualities and R elated B ehavior'.This is a m atter of interpretation.If 'h ate' belongs to this dom ain, does 'love' not also belong here, and why are nXoúaioq and related w ords not included under dom ain 88 (see M alina 1987)?How would native speakers have understood these term s?A nother example might help us a little further.
This is particularly true w ith reg ard to sem an tic do m ain s w hich a re easy to reco g n ise -such as plants, anim als, foods, body and parts of the body and kinship terms.T h ere are, however, num erous fields which are not so clear.A good exam ple is dom ain 25, 'A ttitudes and Em otions', w here we find dyotnow and ócyóaxri grouped together with words such as (|)iXé(o, (|)iXio, (|)iXo6eX(|)ia, (j)iXó8eX(tio<;, d ax o p yo q , ^tiXck»), axevox(upéojj,ai é u xoli; CTnXáyxvoiq, au^inaSéu) ktX.W h at is rem ark ab le, is th a t we do n o t find )aiCT€(iJ here.W e find )i.iaéa)u n d er the subdom ain 'H ate, H ateful' with words such as KOKÍaS niKp'ux*', niKpaii;o)aai, áaxupyéo), GeooTuyfii; KtX.