Examining the effectiveness of ‘Bilingual Dictionary Plus’ - a dictionary for production in a foreign language

EXAMINING THE EFFECTIVENESSOF ‘BILINGUAL DICTIONARY PLUS’ –A DICTIONARY FOR PRODUCTIONIN A FOREIGN LANGUAGE

Batia Laufer and Tamar Levitzky-Aviad:University of Haifa, Israel([email protected])

Abstract

The study investigated the usefulness of a Hebrew-English-English (L1-L2-L2)

mini-dictionary, for production in a foreign language. The dictionary entries included

L1 (Hebrew) words, their L2 (English) translation options, usage specifications,

semantically related English words, and additional L2 meanings for each transla-

tion option. Seventy five students translated thirty six sentences from Hebrew

into English using four dictionaries, one dictionary for nine sentences: a Hebrew-

English-English electronic dictionary, a Hebrew-English-English paper dictionary,

an English-English-Hebrew bilingualized dictionary, and a Hebrew-English bilingual

dictionary. Each sentence contained one target word, defined as ‘simple’ or ‘complex’.

Learners also rated them in terms of usefulness. All electronic dictionary activity was

recorded in log files. The four dictionaries were compared on the number of correct

translations and on dictionary ratings. Log files were analyzed for lookup preferences.

Results demonstrated the superiority of L1-L2-L2 dictionaries in terms of the

effectiveness for a production task and learners’ preferences.

1. Introduction

Surveys of dictionary use indicate that the majority of foreign language

learners prefer bilingual L2-L1 dictionaries and use them mainly to find the

meaning of unknown foreign (L2) words (Atkins 1985, Piotrowsky 1989).

However, if learners writing in L2 need an L2 word designating a familiar L1

concept, they do not readily turn to an L1-L2 dictionary for help. The reason

for this may lie in a serious limitation of most L1-L2 bilingual dictionaries.

They rarely differentiate between the possible L2 translations of the L1 word,

nor do they provide information regarding the use of each translation option.

Because of these limitations, many teachers often discourage their students

International Journal of Lexicography, Vol. 19 No. 2. Advance access publication 10 April 2006� 2006 Oxford University Press. All rights reserved. For permissions,please email: [email protected]

doi:10.1093/ijl/eck006 135 at U

niversity of Haifa L

ibrary on April 29, 2014

http://ijl.oxfordjournals.org/D

ownloaded from

https://ssl.haifa.ac.il/,DanaInfo=ijl.oxfordjournals.org+

from using bilingual dictionaries for writing ( Tomaszczyk 1983, Rundell 1999).

However, if this kind of information, which is essential for using new words,

were added, the L1-L2 bilingual dictionaries would enhance rather than

hinder foreign language production. With this aim in mind, Laufer (1995)

suggested to create a special L1-L2-L2 dictionary for production. An entry in

such a dictionary would include four elements:

(1) L1-L2 translations.

(2) L2 information (definitions, examples, etc.) about each translation option.

(3) Thesaurus-like information, i.e. semantically related words to each

translation option.

(4) Additional L1 meanings of the L2 translations.

These requirements can be fulfilled by an electronic dictionary since such a

dictionary can combine the features of an L2-L1bilingual dictionary, an L1-L2

bilingual dictionary and an L2 monolingual dictionary (Nakamoto 1995,

Nesi and Leech 1999, Kernerman 2000, Tono 2000). For the purposes of

this study, a Hebrew-English-English mini-dictionary for production was

designed following Laufer’s (1995) proposal, and an empirical investigation

was carried out into the usefulness of such a dictionary for second language

writing.

2. Background

2.1 Dictionaries and learner preferences

Learners often admit what lexicographers and language teachers have known

for a long time: that monolingual dictionaries (MDs) provide more detailed

and precise information about the words looked up than bilingual dictionaries

(BDs) (Bejoint 1981, Tomaszczyk 1983, Piotrowski 1989, Marello 1998, Nesi

2002). And yet surveys of dictionary use have shown that learners own, use

and prefer their L2-L1 BDs more than MDs (MacFarquhar and Richards

1983, Atkins 1985, Piotrowski 1989, Scholfield 1999).

One cause for this preference could probably be the lack of direct reference

to L1 in the MD. However good the explanation, or illustrations in L2 might

be, people feel more confident if they can relate the meaning of a foreign word

to a specific lexical entry in their L1 (Tomaszczyk 1983, Snell-Hornby 1987,

Laufer 1995, Cubillo 2002). Second, if the words used in a definition are

unclear, an additional search for explanations of these words may be required.

The L2-L1 BD, on the other hand, relates the meaning of an unknown foreign

word to an L1 word and does not burden the learner with additional unknown

words. Another related cause for this preference of the L2-L1 BD could

perhaps be the users’ desire and need to gain immediate understanding of the

136 Batia Laufer and Tamar Levitzky-Aviad at U




ownloaded from


material they are reading, rather than to learn about the looked up words for

future benefits (Aust, Kelley and Roby 1993). This is particularly true if the

learners are not very proficient, or unskilled in dictionary use. Taking a

pragmatic approach to dictionary use, Piotrowsky (1989: 81) claims that ‘users

will not be helped when they are advised not to use BDs at all. The users will

be helped only when bilingual dictionaries are significantly improved’.

A serious attempt to improve the learners’ dictionaries was made through a

bilingualized dictionary. This dictionary combines the information provided in

the learner’s MD, i.e. definitions, examples of use and grammatical properties

of the words and L1 equivalents as does a BD. The value of the bilingualised

dictionary has been shown by Hartmann (1994), Laufer and Hadar (1997),

Laufer and Hill (2000). By using interviews with informants and direct

observations during a reading task, Hartmann revealed that users from four

different L2 proficiency levels considered the juxtaposition of L2 definitions

and L1 translation equivalents helpful for their comprehension tasks. Laufer

and Hadar (1997) showed that bilingualised entries yielded better results in

comprehending the looked up words than did bilingual and monolingual

dictionaries. The same was true for writing original sentences with these words

except in the case of weak learners. Laufer and Hill (2000) found that different

people had different lookup preferences and that the use of multiple dictionary

information reinforced retention of the looked up words. The dictionaries

which best cater to individual preferences are bilingualised dictionaries as they

provide both monolingual and bilingual information.

2.2 Dictionaries and production in a foreign language

The studies above suggest that the strength of the bilingualised dictionary lies

in its catering for a variety of look up possibilities and personal preferences,

and in combining the monolingual and bilingual information. Its weakness, on

the other hand, lies in unidirectionality, from the target language, e.g. English,

to the mother tongue, e.g. Hebrew. Since the point of reference in such a

dictionary is L2, the learner needs to be familiar at least with an L2 word form

in order to look it up and find the necessary information about it. This means

that the dictionary is inadequate for production in L2, when the learner needs

to find an unfamiliar L2 word for a familiar L1 concept. Even the Longman

Essential Activator (Rundell 1995) which was constructed with L2 production

in mind has serious limitations since it makes an assumption about the user

which may not be true. The assumption is that though the users may not

possess the precise word needed, they are nevertheless familiar with other

words in the same semantic area. They will therefore open the Activator at the

entry of one of these words, check the other words in the semantic field and

select the most suitable one for that purpose.

Examining the Effectiveness of ‘Bilingual Dictionary Plus’ 137 at U




ownloaded from


However, in the process of formulating the thought in the foreign language,

the learner will often get stuck for a word. What would most probably come to

mind is the L1 word that is needed rather than a synonym, antonym, or any

other semantically related word in L2. This is so since words in our dominant

language are more easily accessible than words in languages less familiar to us.

The learner would need the translation options of the L1 concept together with

the grammatical, semantic, and pragmatic specifications of each option and

examples of its use. Since translations into L2 are provided by BDs while

syntactic and semantic information along with examples are provided in the

MD, the learner will have to use both dictionaries for successful production.

Such a procedure is rather tedious and not user-friendly. It is based on the

assumption that the dictionary user owns at least two dictionaries. It is also

time-consuming and would probably be more suitable for particularly

motivated students. Moreover, the users may not always know which of the

L2 equivalents they need to look up in the MD and may end up looking up all

equivalents until finding the right one, thus increasing the lookup time even

more (Axelsen 1993, Rundell 1999).

A slightly more convenient way of accessing the necessary information

would be an L1-L2 index at the end of the MD, e.g. all the Hebrew translations

of the English monolingual entries with page numbers on which the entry can

be found. This indexing was implemented in many bilingualized dictionaries

for beginners and intermediate learners for L1 speakers of Bulgarian, Chinese,

Czech, Hebrew, Italian, Latvian, etc. (produced by KDictionaries Ltd).

The basic disadvantage of such an index, however, is that many L1 words in

the index may have more than one English equivalent. Thus, although the

learners do not need more than one dictionary, they may still need to look

through various English equivalents within the dictionary itself until the most

appropriate one is found. Besides, the Hebrew words included in such a

dictionary are only direct translations of the English words, and are not

necessarily the most frequent Hebrew words, or words for which there is no

simple translation, let alone multiword units.

The need for a special dictionary for production was realized as early as

1950 by Shcherba (in Jarosova 2000), who stated that the construction of

a productive (Russian-French) dictionary had to be based on the so-called

‘translation principle’. This principle means primarily two things. First, the

dictionary should not give an explanation of the L1 (Russian) meaning, but a

real translation into the L2 (French) equivalent, preferably inserted within

a phrase. Second, in cases where the L1 word form has more than one L2

equivalent, the dictionary should show how and when to use which translation.

These principles were echoed almost 40 to 50 years later by Bogaards 1999,

Bogaards and Hannay 2004, Piotrowsky 1989, Scholfield 1999 who pointed out

that unless some additional information on each translation equivalent was

provided, the user might be unable to choose the most appropriate option,





ownloaded from


and by Tomaszczyk (1983) and Snell-Hornby (1987), who suggested a special

BD for productive purposes. Laufer (1995) too proposed to construct such a

dictionary and specified its characteristics. At the time she referred to it as

‘semi-bilingual dictionary for productive purposes’. As this dictionary will be

the object of our study, we will describe it in detail. The term ‘semi’ means half,

and ‘semi bilingual’ may be associated with ‘less than bilingual’. Since the

proposed dictionary includes more information than a standard BD, we will,

from now on, refer to this dictionary as ‘bilingual dictionary plus’.

2.3 A bilingualdictionary for production:‘Bilingual Dictionary Plus’

Laufer (1995) suggested that an entry in a ‘bilingual dictionary plus’ should

include four elements:

(1) L1-L2 translation – The dictionary proposed here starts with an L1 word

which is first translated into L2. Some L1 lookedup words will have several

equivalents in English (the Hebrew word SHIR, for example, is either a

‘poem’ or a ‘song’). In such cases, all the translations will be provided.

(2) Information about the L2 translation options – Information about the L2

translation will consist of the words’ phonological, grammatical and semantic

specifications (all provided in L2), followed by a definition and examples of

use. In other words, Parts 1 and 2 of the entry in our dictionary provide a

mirror picture of the entry found in the existing bilingualized dictionaries – the

English word with its specifications, definition, examples and translation into L1.

(3) Semantically related words – This part resembles a thesaurus. Words

semantically related to the English equivalent of the lookedup L1 word will

be listed with their definitions and L1 translations. The advantage of this

component is in providing the user with an opportunity to select the most

suitable word out of several words in the semantic area.

(4) Additional meanings of the L2 translations – Additional meanings of the L2

translations will be provided in those entries where the English translation

is polysemous or homonymous while the L1 word is not. Here is an example

of such a case. Suppose the learner has looked up the Hebrew word

MOFSHAT, which in English is ‘abstract’ (the opposite of ‘concrete’). But

‘abstract’ has also an additional meaning of ‘summary’. This meaning has

nothing to do withMOFSHAT. This additional meaning could be translated

and illustrated. The advantage of doing this is in preventing the learner from

assuming that each time ‘abstract’ appears, it will mean MOFSHAT.1

We shall now illustrate two types of entries with Hebrew-English examples.

The Hebrew words are transliterated in brackets. In example 1, the English

translation of the Hebrew lookedup word (BERER) has additional meanings

which are presented at the end of the entry. In example 2, the lookedup word





ownloaded from


(NO’ACH) has several English translations. Since these translations in all

their shades of meanings are covered by the Hebrew word NO’ACH, there is

no need to present additional meanings of these words.

Example 1

rrb (BERER)

find out (vt, past tense, past participle found out)

to learn something by study or inquiry: Find out the cost and let me know.

Please find out when the next train leaves.

Related words:

get at (v prep, infml) to manage to find out hlyg (GILA)

I’m afraid we just can’t get at the information; no one will help us.

determine (vt, fml) to find out exactly: ubqw b?yj (CHISHEV VE-KAVA)

The police wanted to determine all the facts/what happened.

detect (vt, fml & tech) to find out: !yjbh (HIVCHIN)

We have been able to detect some improvement as a result of the medicine.

Other meanings of find out:

You’ve broken the vase and if your mother finds out she’ll be angry.rk lu udy(YADA AL KACH)

Don’t steal pens; if you’re found out there’ll be trouble. spt (TAFAS)

Example 2

jwn (NO’ACH)

(1) Comfortable (adj. opp. – uncomfortable) tyzyp jwn (NO’ACH FIZIT)

(a) pleasant to be in or on: a comfortable chair.

(b) free from pain, anxiety, grief: she feels comfortable after the operation.

(2) Convenient (adj. opp. – inconvenient) hm@th, rwdys tnyjbm, jwn(NO’ACH Me-BCHINAT SIDUR)

suitable, that avoids trouble or difficulty, easy to get to:Will it be convenient for

you to start work tomorrow? This is a convenient method of payment. The car is

parked in a convenient place.

Related words: (related to 1b)

tranquil (adj) calm, quiet and peaceful: wl?, ~q? (SHAKET, SHALEV)

She leads a tranquil life in the country.

relaxed (adj) calm and peaceful in body and mind: uwgr (RAGU’A)

He spoke in a relaxed way to his friends.

Lately, several BDs for production have seen light. The First Active Bilingual

Dictionary for Polish Learners of English – the Longman Slownik wspolczesny





ownloaded from


angielsko-polski/polsko-angielski (Polish-English-English dictionary) (Fisiak

et al. 2004) provides the examples in the Polish-English part in the dictionary’s

target language: English. The authors justify this innovation by the fact that

‘Poles do not need to be told how Polish words are used, or at least, that it is

not the kind of information for which they need a dictionary of English. It thus

seemed to make more sense to provide them with English sentences and

sentence fragments illustrating the various translations of the individual senses

of Polish headwords . . . . Most importantly, learners can use the examples as

models for their own production in English.’ (Adamska-Salaciak 2005,

retrieved from http://www.kdictionaries.com). Le Robert & Collins Junior

Bilingue First in English dictionary (Back 2002) is a French-English, English-

French dictionary. But its encoding part (French-English) provides example

sentences of the L2 translation of the L1 headword that illustrate various

points relating to the English (L2) equivalent of the French (L1) headword,

such as prepositional collocations of verbs, þ/� count feature of nouns, the use

of articles, etc. In a separate column, the translation of the headword appears

as used in the example, followed by notes in French that draw attention

to specific points (register, syntax, pronunciation and so on). Back (2005)

illustrates the superiority of this dictionary over ‘conventional’ BDs:

‘In standard dictionaries, for example, French numerals (deux, quinze) are

translated by English numerals (two, fifteen), and the reader is expected to

know that where dates are concerned English uses ordinals (the second, the

sixteenth). In the Junior, this is made explicit via example sentences and an

accompanying note . . .’ (Retrieved from http://www.kdictionaries.com)

With the advent of electronic dictionaries (EDs), the ‘bilingual dictionary

plus’ became easy to construct and simple to use. Whether the ED is installed

on a personal computer, or whether it takes the form of a pocket dictionary,

the principles of storage and retrieval of information are similar to one another

and different from paper dictionaries.

A single electronic dictionary may include a number of functions which

would normally occur in several paper dictionaries. An ED can be used as an

L2-L1 BD, an L1-L2 BD, an L2 MD, and therefore aid both comprehension

and production (Nakamoto 1995, Nesi and Leech 1999: 303, Kernerman 2000,

Tono 2000). For example, the learners can look up an L1 word, i.e. use the

L1-L2 function. Having found the translation, they can then use hyperlinks

in order to explore its specification of use, i.e. shift to the L2 monolingual

function. Furthermore, the ED, particularly computer based, can link the

looked up words to multimedia files (e.g. sounds and pictures) and other

software. Unlike paper dictionaries, EDs can store enormous amounts

of electronic data in a tiny space, including all possible information on the

headwords, and yet the users will not suffer from information overload as

they do not have to access it all. Paper dictionaries, on the other hand do

‘suffer from problems of information overload: too many different types of





ownloaded from

https://ssl.haifa.ac.il/,DanaInfo=www.kdictionaries.com+

https://ssl.haifa.ac.il/,DanaInfo=www.kdictionaries.com+


information in the same dictionary can clutter the page and distract the reader

from the information quest in hand’ (Nesi and Leech 1999: 302).

The use of an electronic dictionary together with built in log files as an

unobtrusive research instrument introduced a rigorous methodology in

investigating learners’ look up behavior and its effect on the comprehension,

production and acquisition of the looked up words. Log files can keep track of

words which have been looked up, type of dictionary information (definition,

translation, example, etc.) that has been selected, the number of times each

word was looked up, and the time necessary for task completion.

Research on EDs from the user’s perspective has focused on the way these

dictionaries are used, and their usefulness as on-line helping tools and as

contributors to incidental vocabulary learning. Some studies compared the

effect of different types of glosses (paper, electronic textual, electronic pictorial,

electronic and video) on reading comprehension, translation, on the number

of words looked up by the learners, the length of time on task and the effect of

gloss type on the reported satisfaction of dictionary users (Leffa 1992, Aust,

Kelly and Roby 1993, Lomicka 1998, Nesi 1999). Others investigated incidental

vocabulary learning via computer glosses (Lyman-Hager et al. 1993, Chun and

Plass 1996, Plass et al. 1998, Laufer 2000, Laufer and Hill 2000, Ajzawa 2002,

Hill and Laufer 2003, Tono 2003).

To our knowledge, no research has been carried out on the usefulness of

‘bilingual plus’ dictionary type on learners’ written production. Such an

investigation is the aim of this paper.

2.4 ‘Bilingual plus’mini-dictionaryin the study

Two versions of ‘bilingual plus’ (henceforth BDþ) mini-dictionary were

constructed for the purpose of the study: a paper version and an electronic

version. The languages used in this mini dictionary were Hebrew and English.

The mini-dictionary had 44 entries for 44 words: 36 entries contained the

target words selected for investigation, and the other 8 entries served as

distracters.

Each entry was constructed according to the principles in Laufer (1995),

as specified earlier in section 2.3. The Hebrew word was provided with its

English translation equivalents and also with semantic and grammatical

information about each of these equivalents, written in English: prepositions

used with the target verbs, an English definition of the Hebrew word,

example(s) of usage of each translation equivalent, possible fixed expressions

with the English words, semantically related words in English, and some

additional meanings of the English translations translated into Hebrew where

the English translated word was polysemous or homonymous while the L1

word was not. The Computerized version was a computer program where the





ownloaded from


entries were organized according to the same principles as the paper version,

with one difference: in the paper version, the entire entry is presented

simultaneously, while in the computerized version, each type of information is

only provided upon the participant’s request.

Figures 1 and 2 show two screens taken from the computerized version of the

BLDþ. Figure 1 shows the list of Hebrew words viewed by the participants

immediately on entering the program. Figure 2 demonstrates the choices

Figure 1: The 44 Hebrew entries as presented in the computerized BDþ

Figure 2: Definitions and usage examples of an English translation equivalent





ownloaded from


made by one of the participants: The participant chose to click on the

Hebrew word hkyz (zika). He then clicked on one of the English equivalents

of that word (credit) and asked for the English definition of this latter word.

Finally, he also asked for examples of usage of the word credit.

3. The Study

3.1 Research questions

We asked the following research questions:

(1) Is there a difference in the usefulness of the following dictionaries for

correct word usage:

(a) a paper bilingualized (L1-L1-L2) dictionary(b) an L1-L2 paper bilingual dictionary(c) an L1-L2-L2 paper bilingual dictionary plus(d) a computerized bilingual dictionary plus?

(2) Do students prefer some dictionaries over others for written production?

(3) Which lexical information of the dictionary entries do students look up for

written production (in a computerized dictionary)?

3.2 Participants

Seventy-five students that represent four English proficiency levels participated

in the study. Sixteen participants of the highest proficiency level were English

majors from the department of English Language and Literature at Haifa

University. Nineteen were university students from other departments (i.e. non-

English majors) who were exempted from courses of English as a Foreign

Language. Seventeen were university students from other departments who

were not good enough for exemption and were required to take two semesters

of English as a foreign language and were, at the time of the experiment,

in their second semester. Twenty three participants were 11th grade high school

students. All students had native or near native competence of Hebrew.

All reported being familiar with the computer environment and knew how to

use the mouse.

3.3 Materials

3.3.1 Dictionary types.Four dictionaries were compared.

(1) The Oxford Student’s Dictionary for Hebrew Speakers (Reif 1993) – An

English-English-Hebrew bilingualized dictionary. This is the dictionary





ownloaded from


approved by the Israeli ministry of education for use in schools and in the

matriculation exams. Even though we believe this dictionary is not very

useful for production tasks, we included it in our study since learners sitting

for school exams are allowed to use this dictionary only.

(2) The Alkalay Hebrew – English Dictionary (Alkalay 2000). Pages including

the entries of the target Hebrew words were photocopied from this BD and

handed out to the participants.

(3) A ‘bilingual plus’ Hebrew-English-English ‘mini’ dictionary (paper

version), specifically written for the purpose of the study, including 44

Hebrew main entries, 36 of which contained the target words.

(4) A ‘bilingual plus’ Hebrew-English-English ‘mini’ dictionary (computerized

version) – a computer program specifically designed for the study. As

mentioned earlier, this program included the same 44 entries as in the

paper version and followed the same principles, with one difference: the

paper version displayed the entire entry, while in the computerized version,

each type of information was provided only upon the participant’s request.

3.3.2 Target words. Two ‘types’ of words were investigated in the study. They

were defined as simple and complex. ‘Simple’ words were Hebrew words with

a one-to-one correspondence of meaning with their English equivalent(s),

e.g. tynlk (kalanit) – anemone; tqcm (matzeket) – ladle or dipper). ‘Complex’

words were of two kinds: (a) Hebrew words which had several meaning

equivalents in English, e.g. hnmzh (hazmana) – ‘invitation, order, indent,

booking, reservation, summons, subpoena’; or (b) Hebrew words different

from their English equivalents in the prepositions they combine with,

e.g lu. h@tmh (hamcha’ al) (literally ‘a check on’) – a check for. . .; jtyn(niteax – transitive verb) – operate on . . ., ‘dissect, analyze’. Thus, complexity

could be semantic or syntactic, or both.

To identify words that were unfamiliar or problematic for all the

participants, a pilot study was conducted with 12 students who would not

later participate in the main study. They represented the four proficiency levels,

3 students from each level. They were asked to translate 48 Hebrew sentences

into English without the use of a dictionary. Each sentence contained a

potential target word for investigation. Following the pilot, only 36 of the

48 words were chosen as target words for the main study. Twelve were nouns,

twelve verbs and twelve adjectives. Twelve of the thirty six words were simple

and twenty four were complex.

3.3.3 Students’ worksheets. Four students’ worksheets were prepared for the

purpose of this research, one worksheet per dictionary type, as will be described

in section 3.3.4. Each worksheet included 9 Hebrew sentences (out of the total

of 36), each sentence had one target word which was underlined. Of the nine

sentences in each worksheet, three included ‘simple’ target words (as defined





ownloaded from


for the purpose of the study) and the other six sentences included ‘complex’

words. Three of the target words in each worksheet were verbs, three were

nouns and three were adjectives. To prevent any unnecessary translation

difficulties, other than those encountered while trying to translate the target

words, the rest of the vocabulary used in the sentences was easy to translate.2

Here are two examples of Hebrew sentences and their translations. The first

sentence includes a ‘simple’ word, the second – a ‘complex’ word:

(aragti shlosha shtichim) <wyh <yjy~? h?l? ytgr@ .1

I wove three carpets today

uwdm udwy yny@w ~p?mh tybl hnmzh ytlbq .2

(kibalti hazmana le-beit mishpat ve-eini jodeja madua)

I got a subpoena and I don’t know why.

3.3.4 Procedure. We asked the participants to translate 36 sentences with the

target words spread over the four worksheets. Each worksheet with nine

Hebrew sentences represented one dictionary type. Thus, for example, one

student may have first been required to use the English-English-Hebrew

bilingualized dictionary with worksheet 1, then proceed to the next condition

and use the Hebrew English BD with worksheet 2, then use the Paper BDþ

with worksheet 3 and finally the Computerized BDþ with worksheet 4. The

order of the worksheets, as well as the dictionary required for each worksheet

changed from one participant to another. Thus there were four dictionary

conditions which were compared. By using different combinations of

dictionary type and worksheet, we made certain that all participants used all

four dictionaries, and each target word was translated with the help of the four

dictionaries though by different students.

On completion of their writing tasks, students were given a preferences

questionnaire asking them to rate, on a scale from 1 to 4, the four dictionaries

from the most useful to the least useful and to justify their rating.

3.3.5 Scoringand data analysis. We carried out a within-subject comparison of

translating different words with different dictionaries. The usefulness of a

dictionary (our concern in the 1st research question) was measured by the

accuracy of English translations chosen by the participants from the dictionary.

Each word translation could receive a score of 0, 0.5, or 1 determined as

follows: 1 point was given only when the most appropriate English equivalent

was provided, when it was spelled correctly and when the right proper

preposition was chosen (where needed). 0 points was given when the English

translation was incorrect or missing, thus creating a semantic mistake, or when

(a needed) preposition was incorrect or missing thus creating a syntactic

error. 0.5 points were given in three cases: when the English translation was





ownloaded from


a possible one, but not the most appropriate one, thus creating a semantic

inaccuracy; when the preposition used was a possible one, but not the best one

presented in the dictionary, thus creating a syntactic inaccuracy; when a

spelling mistake has occurred.

The individual word scores were added up and a total score (percentage out

of the total of 9) was received for each worksheet, i.e. for each dictionary

condition. In addition to the total score, a separate score was calculated for

all the simple words (percentage of the total of 3 points), and a separate score

for all the complex words (percentage of the total of 6 points). A Repeated

Measure Analysis with post-hoc-tests was carried out for each of the

3 comparisons (translations of all the sentences, translations of sentences

with simple words, translations of sentences with complex words).

The learners’ preferences of the dictionaries (research question 2) were

determined by their ratings of each dictionary (from 1 to 4) and their comments

which justified the ratings.

Participants’ preferences of the lexical information (our third research

question) were collected from their computer log files. Each participant first

clicked on a Hebrew word to look up its English translation equivalent(s).

Next, by clicking on one of these translations, at least two options became

available: English definition(s) of the specific English translation, and

example(s) of its usage. Three additional options were sometimes available

(depending on the word): English expressions with that translation equivalent,

its semantically related words and additional meanings of the English

equivalent. The various combinations of look up options that were used by

the participants were grouped into eight major look up patterns and for each

lookup pattern, we calculated the percentage of participants who used it.

4. Results

The following abbreviations will be used in all tables:

BLD the Oxford English-English-Hebrew Bilingualized dictionary

BD the Alcalay Hebrew-English Bilingual dictionary

P-BD þ the Hebrew-English-English paper bilingual dictionary plus

C-BD þ the Hebrew-English-English computerized bilingual dictionary plus.

4.1 The usefulness ofthe various dictionaries for production

Our first research question was: Is there a difference in the usefulness of the

following dictionaries for correct word usage: (a) a paper BLD, (b) an L1-L2

BD, (c) an L1-L2-L2 paper BDþ, (d) a computerized BDþ?

As stated earlier, the dictionaries were compared on their usefulness for all

target words and separately for the translation of ‘simple’ and ‘complex’ words.





ownloaded from


Tables 1–3 present the results in percentage for each of the above comparisons.

Tables 1a–3a display the descriptive statistics, while table 1b–3b show which

differences in means between all the different pairs of dictionaries are

statistically significant.

Table 1b shows that the mean scores of all the responses obtained with the

use of the P-BDþ or with the C-BDþ were significantly higher than those

obtained with the use of the BLD and the BD though not different from each

other. The mean scores obtained with the use of the BD were significantly

higher than those obtained with the BLD.

Table 2b shows that the mean scores for the ‘simple’ words obtained with

the BLD were significantly lower than those obtained with other three

dictionaries. The mean scores obtained with the P-BDþ and C-BDþ were not

different from one another, but were significantly higher than those obtained

with the other two dictionaries.

Table 3b shows that the mean scores for the ‘complex’ words obtained with

the use of the BLD and the BD were not different from each other, but were

significantly lower than those obtained with the P-BDþ and the C-BDþ. As in

the previous two comparisons, no significant differences were found between

the paper and computer versions of the bilingual plus dictionaries.3

4.2 Participants’preferences ofdictionaries

Our second research question was: Do students prefer some types of dictionaries

over others for written production? Tables 4a and 4b present the mean scores

Table 1a: Correct translation scores – all target words

Dictionary type N Min (%) Max (%) Mean (%) SD

BLD 72 0 72 27 19

BD 75 17 89 51 16

P-BDþ 73 39 100 80 18

C-BDþ 74 39 100 80 17

Table 1b: Correct translation scores – differences between means

Dictionary type BLD BD P-BD

BLD

BD 24**

P-BDþ 53** 29**

C-BDþ 53** 29** 0

**Significant at p5 0.01.





ownloaded from


of the participants’ subjective ratings of the four dictionary types. Table 4a

presents the mean scores; table 4b presents the differences between the means

and shows which differences in means between all the different pairs

of dictionaries are statistically significant.

Table 2a: Correct translation scores – simple words


BLD 72 0 100 33 33

BD 75 0 100 89 20

P-BDþ 73 42 100 98 9

C-BDþ 74 67 100 97 8

Table 2b: Correct translation scores: simple words – differences between means


BLD

BD 56**

P-BDþ 65** 9*

C-BDþ 64** 8* 1

*Significant at p5 0.05; **Significant at p5 0.01.

Table 3a: Correct translation scores – complex words


BLD 72 0 100 25 19

BD 75 0 83 31 20

P-BDþ 73 8 100 71 26

C-BDþ 74 17 100 69 26

Table 3b: Correct translation scores: complex words – differences between

means

Dictionary Type BLD BD P-BD

BLD

BD 6

P-BDþ 46** 40**

C-BDþ 44** 38** 2






ownloaded from


A Repeated Measure Analysis with post-hoc-tests showed that the ratings of

P-BDþ and C-BDþ were significantly higher than the ratings of the BLD or

BD. The BD was also rated significantly higher than the BLD. As for the two

BDþ dictionaries, the C-BDþ was rated significantly higher than P-BDþ.

4.3 Participants’ lookup preferences ofdictionaryinformation (in the C-BDþ)

The third research question was: Which parts of the entries do students look

up for written production (in a computerized dictionary)? Figure 3 presents the

frequencies of the participants’ lookups in the computerized dictionary as

recorded on the log files during their work with this dictionary. The figures on

the vertical axis represent the percentage of participants (of 100% participants)

who chose a particular lookup option.4

As mentioned earlier, in 3.3.5, eight major lookup patterns emerged from the

analysis of the computer log files. For all the participants, the first step in each

of these patterns was looking up the English translation equivalent(s). In the

case of some words, some of the participants remained satisfied with

the English translation and did not ask for any further information. Hence,

one major lookup pattern was ‘translation only’ which is listed as pattern 8

below. The other seven major lookup patterns combined the English

translation and additional lexical information in English as follows:

(1) Translation and definition(s) of one or more of these translations.

(2) Translation and example(s) of usage.

Table 4a: Ratings of the four dictionary types

Dictionary Type N Min (of 4) Max (of 4) Mean (of 4) SD

BLD 71 1 3 1.18 0.46

BD 70 1 4 2.13 0.66

P-BDþ 71 1 4 3.13 0.66

C-BDþ 72 2 4 3.65 0.63

Table 4b: Ratings of the four dictionary types – differences between means


BLD

BD 0.95**

P-BDþ 1.95** 1**

C-BDþ 2.47** 1.52** 0.52**






ownloaded from


(3) Translation, definition(s) and example(s).

(4) Translation and thesaurus.5

(5) Translation, definition(s) and thesaurus.

(6) Translation, example(s) and thesaurus.

(7) Translation, definition(s), example(s) and thesaurus.

(8) Translation only without any additional information.

The most common lookup pattern was pattern 3 – ‘translationþ definitions

and examples’. The percentage of participants who chose this pattern, out of

the total number of participants, was 69.3. However, not far behind it was

pattern 8, ‘translations only’, i.e., many participants (66.7%). have chosen to

look only at the list of English equivalents without seeking any additional

information.

5. Conclusion

The idea that a BD should help learners’ production is not new. As pointed out

in the literature review section, some lexicographers and language educators

realized the shortcomings of the BDs, and suggested that translations of L1

words should be supplemented with additional information in L2 which is

necessary for the correct selection and use of foreign language vocabulary.

This idea has been implemented lately in several new BDs which claim to have

provided this kind of information. However, to our knowledge, no empirical

studies have been carried out on the effectiveness of this ‘new concept bilingual

dictionary’, which we refer to as Bilingual Dictionary Plus (BDþ).

The study presented in this paper examined how effective such a dictionary

would be for a task in which L2 (English) translations of L1 (Hebrew) words

were looked up and subsequent choices were made regarding a translation

option and its use. We compared translation scores which were obtained using

0

10

20

30

40

50

60

70

% o

f par

ticip

ants

1. Translation+Definitions 2. Translation+Examples

3. Translation+Definitions+Examples 4. Translation+Thesaurus

5. Translation+Definitions+Thesaurus 6. Translation+Examples+Thesaurus

7. Translation+Definitions+Examples+Thesaurus 8. English translations Only

12345678

Figure 3: Participants’ lookup preferences





ownloaded from


the following dictionaries: a conventional bilingual dictionary, a bilingualized

dictionary, BDþ in its paper version, and in its electronic version. We also

looked separately at learners’ performance when they translated ‘simple’

words – words that had one translation equivalent, and ‘complex’ words –

words that had several translation equivalents, and/or words that were

different from their Hebrew counterparts in the use of a preposition. The

results demonstrated the superiority of the BDþ over the other dictionaries.

Both the paper and the electronic versions proved to be effective, yielding 80%

of correct translations altogether; 98% on the simple words and 70% on

the complex words. The results were significantly better than the results

achieved with the other dictionaries. Hence, Shcherba (1950), Bogaards (1999),

Bogaards and Hannay (2004), Piotrowsky (1989), Scholfield (1999),

Tomaszczyk (1983), Snell-Hornby (1987), Laufer (1995) were right in their

claim that there was a need for a special improved bilingual dictionary for

production. Their intuitions regarding the value of such a dictionary were

confirmed by the results of our study.

If we look at learners’ use of the electronic BDþ and their look up

preferences, we can understand why this dictionary is effective. The most

prevalent lookup pattern was translationþdefinitionþ example. Taken all

together, most lookups consisted of translation and some additional

information, information found only in the BDþ. Apparently what made the

dictionary so effective was the combination of the bilingual and monolingual

information which most learners used.

The thesaurus information suggested by Laufer (1995) was, however, not

popular with our learners. One explanation for this may be that thesaurus

information is less necessary than translations, definitions and examples.

Another explanation may be related to the type of task learners performed.

To translate a Hebrew word correctly in a sentence, the participants had to

find the right English equivalent with its semantic and grammatical properties.

In free writing, on the other hand, people often look for a precise shade of

meaning, or for a synonym in order not to repeat the same word. Including

semantically related words in each entry caters to this need. Further research,

with different writing tasks, is needed to examine the value of this component

in a dictionary entry. As for additional meanings of the L2 translations that

Laufer (1995) suggested, these were not necessary for the translation task

either. But if a dictionary is viewed not only as a reference tool, but also as

a vocabulary learning tool, then this component of the entry may serve as

a basis for vocabulary exercises distinguishing between lexicalization of various

concepts in learners’ L1 and the foreign language.

The unique quality of the BDþ was reflected in learners’ reports. They rated

the BDþ as the most helpful dictionary for the writing task they performed. On

the whole learners valued the electronic BDþ more than the paper one.





ownloaded from


Hence the BDþ proved to be the best in terms of dictionary usefulness,

i.e. the effectiveness for a production task, and in terms of dictionary usability,

i.e. learners’ preference. It is our conviction that, in the future, when more

BDsþ are constructed, they will be widely used for writing in a foreign

language.

Notes

1 Giving the additional meanings of L2 equivalents was motivated by the view of a

dictionary as a learning tool, in addition to its being an aid in comprehension or

production. However, as the effectiveness of this feature has not been tested yet,

we may, in the future, find that the inclusion of additional meanings may not be very

helpful.2 Initially we considered listing the Hebrew target words in the worksheets, without

embedding them in sentences, and asking the participants to use the dictionary provided

along with each worksheet in order to produce their own sentences with the English

equivalents of these Hebrew words. While such a strategy would have allowed a greater

degree of free production which is, in fact, the ultimate productive mode, it nevertheless

does not meet the requirements of an empirical study which calls for more controlled

conditions.3 The results in tables 1–3 present the scores of the entire sample. When we divided

the sample into the 4 proficiency groups, the result patterns, i.e. the significance of

differences were identical. Hence we present only the results of the entire sample.4 Each participant could have chosen a different lookup pattern (out of the

8 patterns) for different words. Thus, each participant appears as a ‘data item’ in more

than one column of the diagram. (This explains why the numbers of all the columns add

up to more than 100%). For our purposes, however, it is the separate columns that

matter. Each column shows what percentage of participants (out of 100% participants)

chose a specific lookup option.5 ‘Thesaurus’ here refers to expressions containing the English translation, words

that are semantically related to the English translation, and/or some of its additional

meanings.

References

A. Dictionaries

Alcalay, R. (ed.) 2000. The Complete English-Hebrew Dictionary. Tel-Aviv: Yedioth

Ahronoth/Chemed Books.

Back, M. (ed.) 2002. First in English, dictionnaire francais-anglais anglais-francais. Paris:

Dictionnaires Le Robert.

Fisiak, J., Adamska-Salaciak, A., Jagla, E., Jankowski, M. and Lew R. (eds.) 2004. The

Longman Slownik wspolczesny angielsko-polski / polsko-angielski. London: Longman.

Reif, J.A. (ed.) 1993. The Oxford Student’s Dictionary for Hebrew Speakers (First

Edition 1986.) Jerusalem: Kernerman/Lonnie Kahn.

Rundell, M. (ed.) 1995. The Longman Essential Activator. London: Longman.





ownloaded from


B. Other literature

Adamska-Salaciak, A. 2005. ‘Longman Slownik Wspolczesny Angielsko-Polski,

Polsko-Angielski: the first active bilingual dictionary for polish learners of

English.’ Kernerman Dictionary News, 13.Ajzawa, K. 2002. ‘Comparing the effect of electronic and paper dictionaries on text

comprehension and incidental vocabulary learning through reading.’ Paper presented

at the 13th World Congress of Applied Linguistics, Singapore.

Atkins, B. T. S. 1985. ‘Monolingual and bilingual learners’ dictionaries: A comparison.’

In R. Ilson (ed.), Dictionaries, Lexicography and Language Learning. (ELT

Documents 120) Oxford: Pergamon Press and the British Council, 15–24.Aust, R., Kelley, M. J. and Roby, W. 1993. ‘The use of hyper-reference and conventional

dictionaries.’ Educational Technology Research and Development, 41: 63–73.Axelsen, J. 1993. ‘The Elf and the Ox. Grammatical information in the Danish-English

dictionaries.’ International Journal of Lexicography, 6: 79–88.

Back, M. 2005. ‘Bilingual dictionaries for learners.’ Kernerman Dictionary News 13.Bejoint, H. 1981. ‘The foreign student’s use of monolingual English dictionaries: a study

of language needs and reference skills.’ Applied Linguistics, 2: 207–22.Bogaards, P. 1999. ‘Access structures of learners’ dictionaries.’ In T. Herbst and

K. Popp (eds.), The Perfect Learners’ Dictionary (?) Tubingen: Max Niemeyer

Verlag, 113–30.

Bogaards, P. and Hannay, M. 2004. ‘Towards a new type of bilingual dictionary’.

In G. Williams and S. Vessier (eds.). Proceedings of the 11th EURALEX International

Conference. Lorient, Universite de Bretagne-Sud, 463–74.Campoy Cubillo, C. M. 2002. ‘Dictionary use and dictionary needs of ESP students:

An experimental approach.’ International Journal of Lexicography 15: 206–28.Chun, D. M. and Plass, J. L. 1996. ‘Effects of multimedia annotations on vocabulary

acquisition.’ The Modern Language Journal 80: 183–98.Hartmann, R. R. K. 1994. ‘Bilingualised versions of learners’ dictionaries.’ In H. von

Gert and E. Zofgen (eds.), Fremdsprachen Lehren und Lernen. Gunter Narr Verlag

Tubingen, 206–17.Hill, M.M. and Laufer, B. 2003. ‘Type of task, time-on-task and electronic dictionaries in

incidental vocabulary acquisition.’ International Review of Applied Linguistics 41: 87–106.Jarosova, A. 2000. ‘Problems of semantic subdivisions in bilingual dictionary entries.’

International Journal of Lexicography, 13: 12–28.Kernerman, L. 2000. ‘Pedagogic lexicography: Some recent advances and some ques-

tions about the future’. In U. Heid, S. Evert, E. Lehmann and C. Rohrer (eds.),

Proceedings of the Tenth Euralex International Congress Vol. II (825–829). Stuttgart,

Germany: EURALEX 2000.Laufer, B. 1995. ‘A case for semi-bilingual dictionary for production purposes.’

Kernerman Dictionary News, 3.Laufer, B. 2000. ‘Electronic dictionaries and incidental vocabulary acquisition: does

technology make a difference?’ in U. Heid, S. Evert, E. Lehmann and C. Rohrer

(eds.), Proceedings of the Tenth EURALEX International Congress Vol. II (849–854).

Stuttgart, Germany: EURALEX.

Laufer, B. and Hadar, L. 1997. ‘Assessing the effectiveness of monolingual, bilingual,

and ‘bilingualised’ dictionaries in comprehension and production of new words.’

The Modern Language Journal 81: 189–96.Laufer, B. and Hill, M. 2000. ‘What lexical information do L2 learners select in a call

dictionary and how does it affect word retention?’ Language Learning and Technology

3: 58–76. http://llt.msu.edu/vol3num2/laufer-hill/index.html. (Retrieved Aug. 10, 2005).





ownloaded from

https://ssl.haifa.ac.il/vol3num2/laufer-hill/,DanaInfo=llt.msu.edu+index.html


Leffa, V. 1992. ‘Making foreign language texts comprehensible for beginners:An experiment with an electronic glossary.’ System 21: 63–73.

Lomicka, L. L. 1998. ‘To gloss or not to gloss: An investigation of readingcomprehension online.’ Language Learning and Technology 1: 41–50. http://llt.msu.edu/vol1num2/article2/default.html. (Retrieved July 1, 2005).

Lyman-Hager, M., Davis, J. N., Burnett, J. and Chennault, R. 1993. ‘Une Vie de Boy:Interactive reading in French.’ In F. L. Borchardt and E. M. T. Johnson (eds.),Proceedings of the CALICO 1993 Annual Symposium on ‘Assessment’ (93–97).

Durham, NC: Duke University.MacFarquhar, P. D. and Richards, J. C. 1983. ‘On dictionaries and definitions.’ RELC

Journal 14: 111–24.Marello, C. 1998. ‘Hornby’s bilingualised dictionaries.’ International Journal of

Lexicography, 11: 292–314.Nakamoto, K. 1995. ‘Monolingual or bilingual, that is not the question: the ‘bilingulised’

dictionary.’ Kernerman Dictionary News, 2.

Nesi, H. 2002. ‘A study of dictionary use by international students at a Britishuniversity.’ International Journal of Lexicography 15: 277–305.

Nesi, H. and Leech, G. 1999. ‘Moving towards perfection: the learners’ (electronic)

dictionary of the future.’ In T. Herbst and K. Popp (eds.), The Perfect Learners’Dictionary (?) Tubingen: Max Niemeyer Verlag, 295–306.

Nesi, H. 1999. ‘A user’s guide to electronic dictionaries for language learners.’International Journal of Lexicography, 12: 55–66.

Piotrowsky, T. 1989. ‘Monolingual and bilingual dictionaries: Fundamental differ-ences.’ In M. L. Tickoo (ed.), Learners’ Dictionaries: State of the Art. SEAMEORELC, Singapore.

Plass, J. L., Chun, D. M., Mayer, R. E. and Leutner, D. 1998. ‘Supporting visual andverbal learning preferences in a second language multimedia learning environment.’Journal of Educational Psychology 90: 25–36.

Rundell, M. 1999. ‘Dictionary use in production.’ International Journal of Lexicography12: 35–54.

Scholfield, P. J. 1999. ‘Dictionary use in reception.’ International Journal of

Lexicography 12: 13–35.Shcherba, L. V. 1950. ‘Predislovie ko vtoromu izdaniu.’ In L. V. Shcherba and

M. I. Matuseich (eds.), Russko-Francuzkij Slovar (Russian–English Dictionary, 3rdedition). Moskow: Gosudarstvennoje izdatel’stvo inostranych i nacional’nych

slovarej.Snell-Hornby, M. 1987. ‘Towards a learner’s bilingual dictionary.’ In A. Cowie (ed.),

The Dictionary and the Language Learner (pp. 159–170). Tubingen: Max Niemeyer

Verlag.Tomaszczyk, J. 1983. ‘On bilingual dictionaries: the the case for bilingual dictionaries

for foreign language learners’. In R. R. K. Hartmann (ed.), Lexicography: Principles

and Practice. Academic Press, London, 41–51.Tono, Y. 2000. ‘On the effects of different types of electronic dictionary interfaces on

L2 learners’ reference behaviour in productive/receptive tasks.’ In U. Heid, S. Evert,E. Lehmann and C. Rohrer (eds.), Proceedings of the Tenth EURALEX International

Congress. Stuttgart, Germany: EURALEX, Vol. II: 855–61.Tono, Y. 2003. ‘The effects of information selection functions in electronic dictionaries

and language tasks on L2 vocabulary acquisition.’ Paper presented at the 13th World

Congress of Applied Linguistics, Singapore.





ownloaded from

https://ssl.haifa.ac.il/,DanaInfo=+
