Dictionary-Making Process with Simultaneous Feedback

Gilles-Maurice de Schryver & D.J. Prinsloo

Dictionary-Making Process with Simultaneous
Feedback from the Target Users to the Compilers

Gilles-Maurice DE SCHRYVER and Daan J. PRINSLOO,
Gent, Belgium and Pretoria, South Africa


Abstract
Since dictionaries are ultimately judged by their target users, there is an urgency to provide for the
target users needs. In order to determine such needs more accurately, it has become common practice
to submit users of a dictionary to a series of tests. Unfortunately, in most cases such feedback comes
too late because it can at best be considered for implementation in the revised edition of the dictionary.
In this article it is argued that feedback from the target users should be obtained while compilation is
still in progress. This dictionary-making process is therefore referred to as Simultaneous Feedback
and can be understood as entailing a method in terms of which the release of several small-scale
parallel dictionaries triggers off feedback that is instantly channelled back into the compilation process
of a main dictionary. Following a theoretical presentation of the new concept, its feasibility is
illustrated with reference to the creation of CilubDutch and SepdiEnglish dictionaries.

1. Introduction

Since the lexicographer is constantly striving to enhance the quality of a dictionary, the
active implementation of the target users needs has become all-important in modern-day
lexicography, cf. e.g. [Wiegand 1998] and [Van der Merwe-Fouch 1999]. It is however
not easy for the compiler to determine exactly what these needs are. Recently, dictionary
users have been subjected to various tests meant to monitor what they actually get out of
their dictionaries. [Atkins 1998a] and [Bogaards 1998] are excellent examples of the
sincere efforts currently being made to research dictionary use. The problem however is
that such research is, in general, carried out at a stage when the dictionaries are already
completed and published. Feedback from the target users can therefore only be
implemented in forthcoming editions of these dictionaries. Formulated differently, it means
that the feedback obtained in this way comes too late.

In this article it is argued that feedback from the target users, indispensable in the
compilation of any present-day dictionary, should be obtained while the compilation of the
dictionary is still in progress. This can be achieved by launching several small-scale parallel
dictionaries from which informal as well as formal target user feedback is instantly channelled
back into the compilation process of a main dictionary. Such simultaneous feedback is not
only invaluable in the compilation of the macrostructure but also provides useful information
for improvement on the microstructural level.

2. Theoretical framework of Simultaneous Feedback

A schematic representation of the theoretical framework of Simultaneous Feedback is shown
in Figure 1.

197

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING
Figure 1: Theoretical framework of Simultaneous Feedback

TARGET USERS COMPILERS DICTIONARIES

2
6
Main electronic
corpus:
Word-frequency
studies
& The lemma-
sign list
Main electronic
corpus:
Concordance
lines
& Dictionary
articles
M
A
I
N

D
I
C
T
I
O
N
A
R
Y

based on

M
A
I
N

C
O
R
P
U
S
Formulation of a
theoretical
model
& Preliminary
alysis of targe an t
users' needs
Reanalysis of
target users'
needs
& Formal
academics'
feedback
Continuous
monitoring of
target users'
needs
& Recruiting of
additional staff
4
8
9
5
1
More parallel
dictionaries used
as experimental
tools in order to
refine the main
dictionary
10
Second parallel
dictionary
based on a
second parallel
test-corpus
First parallel
dictionary
based on a
first parallel
test-corpus
3
Informal and
formal feedback
on the
first parallel
dictionary
7
Informal and
formal feedback
on the
second parallel
dictionary
198
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

In Figure 1 one recognises the three primary constituents of any dictionary compilation
process, viz. target users, compilers and dictionaries. The framework itself should be read
from left to right, and from top to bottom [1] through [10]. The compilers central task is
the compilation of a main dictionary depicted by the central arrow in Figure 1. The first
component in Figure 1 is also the first task. That is, the formulation of a theoretically
motivated model for the structure and contents of the dictionary to be compiled, must be
followed by a preliminary analysis of the needs of the potential target users [1]. It is of
paramount importance that, from this initial stage onwards, information concerning these
target users needs be gathered through informal and formal consultations with the future
target users. As such, feedback is simultaneously introduced right from the very start.

Since any modern dictionary is to derive its data from a corpus, the compilers have to build
and query an electronic corpus for the specific language(s) first. As a result, the compilers
cannot start the compilation of the main dictionary the central arrow in Figure 1 right
away and are moreover confronted with the prospect of an extremely time-consuming
undertaking. In order to overcome this deadlock, the main dictionary project is instantly
complemented with a series of small and inexpensive parallel dictionary projects [2], [6],
[10], etc. These parallel projects have basically the same structure, contents and target users
as the main project and are to derive their data from small-scale parallel test-corpora. From
the release of the first parallel dictionary onwards, informal and formal feedback is received
from the parallel projects and channelled back into the time-consuming main project [3],
[7], etc. From that instant, the compilation of the main dictionary becomes a true work in
progress with simultaneous feedback from the target users to the compilers.

The parallel projects are thus used as experimental tools to test a plethora of strategies in
order to refine the presentation of the information in the main project under construction.
Once a structured main corpus has been built, word-frequency studies can be done to assist
the lexicographers in the compilation of the lemma-sign list of the main dictionary [4].
Subsequently, concordance lines, also derived from the main corpus, supplement and
confirm the compilers intuition in the compilation of the main dictionary articles [8].
Until completion of the main project the parallel projects continue to elicit feedback [5],
[9], etc. All this simultaneous feedback ultimately enables the compilers to select the most
appropriate blend of lexicographic procedures to ensure the most effective retrieval of
information by the target users in the main dictionary. Hence, within the framework of
Simultaneous Feedback, we cannot agree with the claim that "Lexicography is not a terrain
in which you can experiment from one day to the next" [Lombard 1994, 211]. On the
contrary, complying with the target users needs means doing just that!

Thus far this theoretical framework, which can be considered as the backbone of any
potential dictionary project, has successfully been applied in compiling a pocket-size
bilingual CilubDutch dictionary, and is currently being used for the creation of a large
SepdiEnglish dictionary.

Due to space restrictions we can unfortunately not expound on every component of the
framework. Instead, for each of the two dictionary projects, we will focus on one very
specific aspect. For the CilubDutch dictionary this aspect will be the lemmatisation of
nouns on the macrostructural level, and for the SepdiEnglish dictionary we shall look
into the microstructural handling of a new grammatical convention.

199

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING

3. Compiling a CilubDutch dictionary
and the concept of Simultaneous Feedback

3.1. Formulation of the problem underlying the Cilub case study

As far as the lexicographical treatment of nouns in Bantu languages is concerned, the
dilemma has always been "How does one successfully lemmatise nouns?" since both
singularity and plurality are indicated by prefixes, cf. [Prinsloo/de Schryver 1999].
According to [Bennett 1986, 4] such "arrangement difficulties, however, are no worse than
those facing the lexicographer in, for example, a Slavic language, where prefixes are
almost as great a problem as in Bantu". Yet, answering the modern-day metalexicographic
call, the arrangement should first and foremost answer the target users needs.

3.2. Collecting feedback for the Cilub case study

When it comes to the macrostructural treatment of nouns, the preliminary analysis seemed
to indicate that the target users preferred nouns to be entered as singulars, followed by the
gender. The two poles of the gender seemed sufficient to enable the users to form, when
applicable, the plural. This arrangement was instantly implemented in a first parallel
dictionary, the Lexicon CilubNederlands [de Schryver/Kabuta 1997], henceforth LCN.
As a case in point the nouns in (1), taken from LCN, can be considered.

(1a) bidy [8/8 -dy] cf syn nshma

(1b) kabji [12/13; dim dibji] papiertje; ~ k bwanga (dokters)voorschrift

(1c) mupy [3/4] 1 adem; ademhaling; 2 geur

Here the genders 8/8, 12/13 and 3/4 were thought to enable the target users to infer the
plural forms bidy, tubji and mipy respectively.

On the day LCN (the first parallel project) was released, a series of Informal Files were
opened and were meant to track every possible remark concerning the lexicon and its use.
As far as dictionary users are concerned, we are convinced that these files represent a rare
firsthand insight into target users (initial) opinions regarding a dictionarys release and its
subsequent use. It should be stressed that all these various opinions were recorded in our
capacity as natural participant observers among learners of Cilub at the Department of
African Languages and Cultures in the University of Ghent.

Besides informal feedback, its formal counterpart was also collected and subsequently
contrasted with the informal feedback. From the manifold standardised methods that are
available to obtain formal feedback we opted for questionnaires through a mail survey. The
Questionnaire itself was four pages long and was packed with fifty questions covering all
the various aspects of LCN, as well as some potential future endeavours. In spite of this
large number of questions with many of them even subdivided into sub-questions nearly
all respondents tried their hands at every single question.

200
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

3.3. Analysing feedback for the Cilub case study

In the Questionnaire, seven questions focused specifically on nouns, of which four dealt
with the lemmatisation of nouns on the macrostructural level. The latter four, with the
answers by each competence level (elementary, intermediate and advanced) expressed in
%, are shown in Figure 2.

In Q23 the target users were first reminded of the lemmatisation approach utilised in LCN
(an approach implemented following the preliminary analysis). We then sought to learn
whether or not this approach (now that it had been implemented and that the target users
had had the chance to familiarise themselves with the outcome) was indeed what the target
users needed. The results revealed that for only 58% of the elementary learners LCNs
approach corresponded to the way they would like to look up nouns. For intermediate and
advanced learners, LCN was on the right track as respectively 93% and 91% of them
marked LCNs approach as conforming to their need. Since 25% of the elementary learners
opted for both singulars and plurals, whilst none of the more advanced learners did, this
suggested a certain unfamiliarity with the noun class system among beginners. The target
users need is thus: For intermediate and advanced learners one can safely lemmatise nouns
under their singulars as long as provision is made for irregular plurals, but for elementary
learners additional guidance on the noun class system will have to be provided for if this is
to be a successful approach. Alternatively, one might even have to consider lemmatising
both singulars and plurals for inexperienced users.

In Q24 we wished to know where the target users want to find irregular plural forms. At
first glance the results seem just a jumble, but when one considers the large number of
respondents who added "under both", a certain logic appears. Moreover, one target user
(rightly) observed that cross-references should link plurals and singulars. Hence, the target
users need is: Irregular plural nouns must both be mentioned within the article of their
singular forms and be lemmatised under their proper alphabetical position, whilst cross-
references should link both forms.

In Q25 we sought to learn whether target users need gender information, and if so, in which
format. What is extremely interesting here are not the options ticked off, but the option not
one single target user chose, namely that class numbers as well as nominal prefixes are
redundant. The results show that roughly half of the elementary and intermediate learners
and nearly all the advanced learners opt for class numbers. The target users need is thus:
Noun gender information is crucial and should, for advanced learners, preferably be coded
using class numbers, whilst elementary and intermediate learners might desire to receive
some additional guidance.

In Q28 we wanted to know the target users opinions about the stem tradition. Their
judgement is extremely scathing concerning this approach, as 67% of the elementary, 87%
of the intermediate and no fewer than 91% of the advanced learners consider this tradition
to be much too complicated. Hence, the target users need is: Under no circumstances
should nouns be lemmatised according to the stem tradition.

These four target users needs (derived from the analysis of the Questionnaire)
corresponded with the data in the Informal Files. It was therefore obvious that we required

201

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING

Figure 2: Q23, Q24, Q25 and Q28 from the LCN Questionnaire, including the stratified
responses

LCN QUESTIONNAIRE RESULTS (in %)
Question Answer Level
suggested by us added by target user elem. int. adv.
23. In the Lexicon nouns are entered under
their singular. Some plurals, however, had
to be provided for. How would you like to
look up nouns?
solely as singular
as singular, with provision for some plurals
always both as singular and plural
other: .
(no answer)
11
58
25
.
6
7
93

.

9
91

.

24. In your view, where should irregular
plural forms be presented?
within the article of their singular
under their plural (as a separate headword)
other: .
under both
(no answer)
28
28
.
39
6
40
20
.
33
7
27
27
.
36
9
25. In the Lexicon the class numbers for
singular and plural are given. Instead of
numbers one could use the nominal
prefixes themselves. According to you,
are:
class numbers handier than nominal prefixes
nominal prefixes handier than class numbers
class numbers & nominal prefixes redundant
other: .
use both, e.g. 1 mu-
(no answer)
58
31

.
6
6
40
60

.


82
9

.
9

28. One could also enter nouns under
their stem. In this way one would for
instance find muntu, bantu, buntu, cintu,
bintu and kantu under -ntu. This would
be:
extremely user-friendly
much too complicated
other: .
the Lexicon would become too large
better: both as stem and noun
interesting to add the stem under the noun
it reveals lexical relations
(no answer)
11
67
.
6
6
6

6
7
87
.



7


91
.



9

202
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

a way to solve a much-desired additional guidance on the noun class system for elementary
learners (and to a lesser extent also for intermediate learners). This requirement was
instantly implemented in a second parallel dictionary, the Beknopt woordenboek Cilub
Nederlands [de Schryver/Kabuta 1998], henceforth BCN. In BCN every page on the left-
hand side displays an inserted text at the bottom which lists the most important affixes, as
shown in (2).

(2) Even Repetitive Inserted Text (ERIT) in BCN

Affix. 1pe 1pm 2pe 2pm 1 2 3 4 5 6 7 8 11 12 13 14 15 16 17 18
NP mu- ba- mu- ba- mu- ba- mu- mi- di- ma- ci- bi- lu- ka- tu- bu- ku- pa- ku- mu-
SC "N- tu- "u- nu- /"- b- "- "- d- "- c- b- l- k- t- b- k- p- k- m-
OC -N- -t- -ku- -n- -mu- -b- -"- -"- -d- -"- -c- -b- -l- -k- -t- -b- -k- -p- -k- -m-
PP u- `b- u- `b- u- `b- `- `- `d- `- `c- `b- `l- `k- `t- `b- `k- `p- `k- `m-
OM c- kat- k- kan- k- kab- ka- ka- kad- ka- kac- kab- kal- kak- kat- kab- kak- kap- kak- kam-

With this repetitive inserted text we intended to provide all the target users with an easy
access to the noun class system (and its grammatical implications). The tabulated affixes
are straightforward. Taking the canonical form of the nominal prefixes for both participants
and classes as a point of departure, the table then lists the subject concords, followed by the
object concords and the pronominal prefixes, and ends with the morphemes used to form
negative tenses. In other words, the 100 tabulated items synthesise the core of the Cilub
concordance system at a single glance.

Just as was the case with the first parallel dictionary, the release of the second parallel
dictionary instantly triggered off feedback. As far as the macrostructural treatment of nouns
is concerned, the analysis of this feedback quickly indicated that virtually all target users
were satisfied with BCNs approach. The main dictionary therefore incorporates this
approach, and as such, one of the long-standing lemmatisation problems in Bantu languages
namely "How does one successfully lemmatise nouns?" was solved by those for
whom the dictionary is intended.

4. Compiling a SepdiEnglish dictionary
and the concept of Simultaneous Feedback

In evaluating the feedback for the Cilub case study in the previous paragraph, the emphasis
was on one specific problematic aspect on the macrostructural level. In this paragraph the
feedback obtained from the target users will be analysed in respect of one particular
microstructural aspect, namely the creation of a digestible convention for communicating a
complicated grammatical structure to the users.

4.1. Formulation of the problem underlying the Sepdi case study

Consider the situation where an inexperienced learner reads a Sepdi novel and is confronted
by the to him/her unknown word thue. Since circumflexes are not indicated in literary works,
a total of eleven possible meanings could be conveyed, as shown in (3).

203

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING

(3a) ga ba thue they are not helping (Indicative)
(3b) ge ba sa thue if/while they are not helping (Situative)
(3c) ba ba sa thuego those who do not help (Relative)
(3d) gore ba thu so that they must help (Subjunctive)
(3e) gore ba se thu so that they must not help (Subjunctive)
(3f) go se thu not to help (Infinitive)
(3g) ba thue they (usually) help (Habitual)
(3h) ba se thue they (usually) do not help (Habitual)
(3i) ba se thu they then did not help (Consecutive)
(3j) ba thu! help them! (Imperative)
(3k) se thu! dont help! (Imperative)

Without proper lexicographical treatment or proper guidance the user will be lost. (S)he wont
even be able to ascertain whether the word has a positive or a negative meaning. The real
challenge is to formulate a convention which will be short and compact, be reasonably user
friendly, enable the users to clearly distinguish between positive and negative meanings, and
act as a bridge for the inexperienced users to the mini-grammar and detailed grammatical
discussion of the verb in grammar books.

Firstly, instances of where the verb ends in - should not only be included in a dictionary but
should also be distinguished from those cases where the verb ends in -e. Secondly, positive
and negative forms with similar endings are to be separated.

For verbs ending in -e (3a, 3b, 3c, 3g and 3h) the convention given in (4) can be designed.

(4) thue help; ~go, ga/sa/se (..) ~ not help

The lexical lemma and its lemmatic address provide for the positive form of the Habitual (3g),
while the second part caters for the multiword lexical items by including and treating various
sublemmata. This section of the article contains the lexical items thuego (3c), ga thue (3a),
sa thue (3b) and se thue (3h) included as grouped lemmata, cf. [Hausmann/Wiegand 1989,
349]. [Wiegand 1989, 391] and [Wolski 1989, 364] categorised sublemmata grouped into an
article headed by another lemma as either niched or nested lemmata. The thue article displays
the occurrence of nested lemmata. This nest includes the sublemma with the relative suffix
-go. That the verb stem can either directly follow the ga, sa or se or be separated from it by
other words is provided for by the brackets and periods.

For verbs ending in - (3d, 3e, 3f, 3i, 3j and 3k) the convention presented in (5) can be
designed on the same principles.

(5) thu(ng)(!) help(!); se (..) ~ not help(!)

Finally, conventions (4) and (5) can even be combined to render a short and compact
convention which sacrifices a little detail in favour of slightly more user-friendliness, as shown
in (6).

(6) thue, thu help; ..ga/sa/se..~ not help

204
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

The first part of this powerful convention (thue, thu help) has a lemmatic address and
conveys the information that a verb ending in -e or - has a positive meaning of must do
something (3d, 3g and 3j). The second part of this convention (..ga/sa/se..~ not help)
introduces the nested lemmata. Their sublemmatic address conveys the information that if a
ga, sa or se occurs before thue or thu, one is dealing with a negative meaning (3a, 3b, 3c,
3e, 3f, 3h, 3i and 3k).

4.2. Collecting feedback for the Sepdi case study

In terms of the point of departure of this article, namely that feedback from the target users
should be obtained while the compilation of a dictionary is still in progress, Exercises and a
Questionnaire were compiled to test, among other aspects, the viability of the ..ga/sa/se..~
convention formulated under (6). The interest of the researchers was focused on the feedback
from a group of beginners/learners of Sepdi. In addition, a group of second language and
mother tongue speakers of Sepdi was also subjected to the Exercises and Questionnaire. Both
groups were a random selection of first-year Sepdi students at the Technikon Pretoria, and
both groups used the same dictionary for all the tests, namely SeDiPro 1.0 [Prinsloo/de
Schryver 2000], a first parallel dictionary SepdiEnglish containing 25,000 lemmata.

4.3. Analysing feedback for the Sepdi case study

Consider (7) as a selection from the phrases testing the viability of the ..ga/sa/se..~ convention,
and taken from the Exercises.

(7a) ga ba thue moruti yo
(not-they-help-reverend-this)
they do not help this reverend

(7b) ba thue, o se ke wa twafa
(them-help, you-not-even-you-be lazy)
help them, you really must not be lazy

All the students correctly concluded that they were dealing with a negative meaning in (7a)
and a positive meaning in (7b). This is significant for the beginners/learners as they had no
other means than (6) at their disposal.

In a second round of Exercises the verb bone was selected, since apart from the range of
positive versus negative meanings in terms of (3), the users had to consider the past tense
form, as well as the meaning fourth. Thus three different articles in SeDiPro 1.0 given as (8).

(8a) bon fourth

(8b) bne have seen; have experienced; ~go which have seen; which have experienced

(8c) bne, bn must see; must look; ..ga/sa/se..~ not see; not look

The to-be-translated phrases in the Exercises were as shown in (9).


205

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING

(9a) ke nyaka gore ba bone taba ye gabotse
(I-want-that-they-see-matter-this-very well)
I want them to understand this matter very well

(9b) gore ba se di bone
(so that-they-not-it-see)
that they must not see it

Although (9a) is a relatively difficult sentence for a beginner/learner to decode, as in the case
of thue above, the majority of beginners/learners managed to arrive at a positive meaning
must see/understand. In the case of (9b), however, almost every student got it wrong with
answers ranging from have seen, to must see and have not seen. A possible reason for this
is that in the case of bone, compared to thue, the user was overwhelmed by too many possible
meanings in the dictionary, cf. (8). In the case of (9a), in contrast to (9b), the user was assisted
by the availability of sufficient context in order to arrive at the correct option.

The observation that beginners/learners are capable of successfully using the ..ga/sa/se..~
convention, was also confirmed by their response to Q17 in the formal SeDiPro 1.0
Questionnaire which concluded the investigation, cf. Figure 3.

Figure 3: Q17 from the SeDiPro 1.0 Questionnaire, including the stratified responses

SeDiPro 1.0 QUESTIONNAIRE RESULTS (in %)
Target group
Question Answer
learner speaker
17. In the dictionary you will find:
phele, phel must live, must stay alive; ..ga/sa/se..~ not
live, not stay alive
Could you explain in your own words what this
convention means?

correct explanation
wrong or no explanation


100


20
80

The results clearly show that all the beginners/learners answered this question correctly, while
as many as 80% of the second language and mother tongue speakers of Sepdi were unable to
answer it, or answered it incorrectly. The latter group also generally failed to answer Q13 and
Q30, as can be seen from Figure 4.

The inability of second language and mother tongue speakers to comprehend standard
dictionary conventions such as /, ~ or = is of great concern to the compilers of Sepdi
dictionaries. Failure to answer these questions can be attributed to a presumed lack of
dictionary culture.
206
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

Figure 4: Q13 and Q30 from the SeDiPro 1.0 Questionnaire, including the stratified
responses

SeDiPro 1.0 QUESTIONNAIRE RESULTS (in %)
Target group
Question Answer
learner speaker
a.
correct word
wrong or no word

86
14


100
13. In many dictionaries the headword is replaced by a
tilde (~) within an article. This is also done in the
dictionary you used. In the dictionary you will find:
ntoma bite me; ~ tsb tell me a secret
a. Which word does the tilde (~) replace here?
b. How do you say tell me a secret in Sepdi?
b.
correct expression
wrong or no expression

71
29

20
80
a. correct suggestion
wrong or no suggestion
43
57
10
90
30. When you see something like this:
feela = fla
a. Do you know what you should do?
b. Do you know why this was done like this?
b. correct suggestion
wrong or no suggestion
43
57

100


From a metalexicographic perspective, this had already been pointed out by e.g. [Gouws 1999,
7, 11], while [Atkins 1998b, 3] observed that "the speakers of African languages have not in
their formative years had access to dictionaries of the richness and complexity of those
currently available for European languages. They have not had the chance to internalize
the structure and objectives of a good dictionary, monolingual, bilingual or trilingual".

The failure of second language and mother tongue speakers of Sepdi to interpret the
..ga/sa/se..~ convention in the first parallel dictionary however does not pose real problems for
the continued utilisation of the convention in the main project. This convention proved to be a
useful if not vital decoding tool for the beginner/learner of Sepdi which is not needed by the
second language or mother tongue speaker since (s)he is familiar with the grammatical rules of
the language. Yet, on a different level, and probably much more important, the ..ga/sa/se..~
feedback triggered the first manifest recording of a serious lack of dictionary culture.
Therefore, while the second parallel dictionary is being compiled, important efforts are
directed towards the explicit and systematic teaching of dictionary skills in the classroom, as
e.g. suggested by [Chi 1998, 566]. Formulated differently, while the main dictionary is being
compiled, the future users are simultaneously being trained in using it, as early feedback
indicated that there is a great need for this.

207

Proceedings of EURALEX 2000
CORPUS-BASED DICTIONARY MAKING

5. Concluding remarks

Simultaneous Feedback is a dictionary-making process which, through the launching of
parallel dictionaries, enables the instant implementation of the target users needs. These
parallel dictionaries moreover answer an urgent desideratum, in that the target users can
instantly be provided with lexica while awaiting an unabridged electronic-corpus based main
dictionary.

The feasibility of this new dictionary-making process was illustrated with two case studies
taken from two different bilingual dictionary projects. The Cilub case study indicated how
two parallel dictionaries enabled the lexicographers to pinpoint and implement the target
users needs as far as a specific macrostructural treatment is concerned. The Sepdi case
study focused on a microstructural topic, and concluded that the suggested convention is a
feasible one. At the same time the latter study brought to the fore that the teaching of
dictionary skills should be considered part and parcel of the dictionary-making process.

We trust that these two small illustrations give an insight into the truly overwhelming
possibilities offered by Simultaneous Feedback. Indeed, simultaneously channelling feedback
from parallel projects into the main project provides lexicographers with the means to
integrate all criticism into the very compilation methodology itself. We are consequently
convinced that the concept of Simultaneous Feedback successfully provides present-day
lexicographers with an inexpensive flying start and a tool for producing dictionaries compiled
according to the latest trends and most modern approaches in lexicography.

References

[Atkins 1998a] B.T. Sue Atkins (Ed.): Using Dictionaries, Studies of dictionary use by language
learners and translators, (Tbingen: Max Niemeyer Verlag) 1998
[Atkins 1998b] B.T. Sue Atkins: Some discussion points arising from Afrilex-Salex98,
(Unpublished course evaluation document, University of Pretoria) 1998
[Bennett 1986] Patrick R. Bennett: "Grammar in the Lexicon: Two Bantu Cases", in: Journal of
African Languages and Linguistics 8/1 1986: 1-30
[Bogaards 1998] Paul Bogaards: "Scanning Long Entries in Learners Dictionaries", in: Thierry
Fontenelle et al. 1998: 555-63
[Chi 1998] Man Lai Amy Chi: "Teaching Dictionary Skills in the Classroom", in: Thierry
Fontenelle et al. 1998: 565-77
[de Schryver/Kabuta 1997] Gilles-Maurice de Schryver and Ngo S. Kabuta: Lexicon Cilub
Nederlands, (Ghent: Recall) 1997
[de Schryver/Kabuta 1998] Gilles-Maurice de Schryver and Ngo S. Kabuta: Beknopt woordenboek
CilubNederlands & Kalombodi-mfndilu k Cilub (Spellingsgids Cilub), (Ghent: Recall)
1998
208
Dictionary-Making Process with Simultaneous Feedback
Gilles-Maurice de Schryver & D.J. Prinsloo

[Fontenelle et al. 1998] Thierry Fontenelle et al. (Eds.): Actes Euralex98 Proceedings, (Lige:
University of Lige) 1998
[Gouws 1999] Rufus H. Gouws: Die maatskaplike gerigtheid van die metaleksikografie in n
meertalige samelewing, (Stellenbosch: Universiteit van Stellenbosch Drukkery) 1999
[Hausmann/Wiegand 1989] Franz J. Hausmann and Herbert E. Wiegand: "Component Parts and
Structures of General Monolingual Dictionaries: A Survey", in: Franz J. Hausmann et al. 1989-
1991: 328-60
[Hausmann et al. 1989-1991] Franz J. Hausmann et al. (Eds.): Wrterbcher / Dictionaries /
Dictionnaires, Ein internationales Handbuch zur Lexikografie / An International Encyclopedia
of Lexicography / Encyclopdie internationale de lexicographie, (Berlin: Walter de Gruyter)
1989-1991
[Lombard 1994] Frikkie J. Lombard: "Lexicographer, Linguist and Dictionary User: An Uneasy
Triangle?", in: Lexikos 4 1994: 204-14
[Prinsloo/de Schryver 1999] D.J. Prinsloo and Gilles-Maurice de Schryver: "The Lemmatization
of Nouns in African Languages with Special Reference to Sepedi and Cilub", in: South African
Journal of African Languages 19/4 1999: 258-75
[Prinsloo/de Schryver 2000] D.J. Prinsloo and Gilles-Maurice de Schryver (Eds.): SeDiPro 1.0,
First Parallel Dictionary SepdiEnglish, (Pretoria: University of Pretoria) 2000
[Van der Merwe-Fouch 1999] Michele Van der Merwe-Fouch: The User Perspective in South
African Lexicography, (Unpublished paper presented at the Fourth International Conference of
the African Association for Lexicography, University of Pretoria) 1999
[Wiegand 1989] Herbert E. Wiegand: "Aspekte der Makrostruktur im allgemeinen einsprachigen
Wrterbuch: alphabetische Anordnungsformen und ihre Probleme", in: Franz J. Hausmann et al.
1989-1991: 371-409
[Wiegand 1998] Herbert E. Wiegand: Wrterbuchforschung: Untersuchungen zur
Wrterbuchbenutzung, zur Theorie, Geschichte, Kritik und Automatisierung der Lexikographie.
1. Teilband, (Berlin: Walter de Gruyter) 1998
[Wolski 1989] Werner Wolski: "Das Lemma und die verschiedenen Lemmatypen", in: Franz J.
Hausmann et al. 1989-1991: 360-71

209