Abstract

The purpose of this paper is to make a contribution to the ongoing call for new ideas in designing electronic dictionaries (e-dictionaries). The paper outlines the design framework of a Greek-English e-dictionary for production in English as a foreign language. To delimit the task of designing the dictionary, which is approached from a linguistic rather than a technical perspective, we first define the properties of the dictionary and its user profile. The main part of the paper describes (a) the elements of the user interface, and (b) the entry components. To illustrate the bilingual entry design, I provide a sample from the entry for πετάω ‘fly’. The proposed design is applicable to any language pair, for which a pre-lexicographic database is available.

Keywords: bilingual lexicography, encoding, user profile, user interface, metalanguage, lexicography and terminology

1. Introduction

This paper focuses on the bilingual dictionary, i.e. the meeting point between translation and lexicography. Although users’ preference for bilingual dictionaries is well documented (Tomaszczyk 1983; Snell-Hornby 1987; Rundell 1999; Hannay 2003; Kernerman 2007; Chon 2009), language teachers have serious reservations about the use of bilingual dictionaries particularly in the context of L2 production (encoding). This context poses a major problem for compilers of print bilingual dictionaries due to spatial and structural limitations (Sharpe 1995; Atkins & Rundell 2008). In order for a dictionary to facilitate encoding, it should help users make appropriate lexical choices and incorporate L2 items into context (Bogaards 1996; Rundell 1999; Chon 2009).

The problem is, however, that whereas entry structure makes information about the headword readily incorporable, it is difficult to provide information on the conditions of use for each translation option. In this respect, the electronic medium seems to open up exciting opportunities for providing access to the diverse information types required in encoding through a single interface. Nevertheless, it is widely observed that current electronic dictionaries are frequently mere copies of traditional paper dictionaries, rather than newly developed products expressly designed for the electronic medium (Sharpe 1995; Nesi 1999; de Schryver 2003; Oppentocht & Schutz 2003; Hallsteindóttir 2007; Pastor & Alcina 2010).

In this light, the paper proposes a novel design framework for presenting lexicographic information in a Greek-English e-dictionary tailored to production needs in English as a foreign language. Throughout the design process we face the challenge of using Modern Greek as the presentation language of the dictionary; an attempt is made to use labels which are neither too long nor too technical or complex, and to offer a translation solution when there is no established terminology in M. Greek.

2. Planning a novel bilingual dictionary

The present work builds on the idea that the compilation of dictionary entries can be systematized by structuring corpus-derived lexical information in a pre-lexicographic database (Atkins & Rundell 2008: 100-101). A product-independent lexical database for English and M. Greek underlies the production of the proposed dictionary (see Dalpanagioti 2011). The analysis and transfer stages of compiling the database may lie outside the scope of this paper; yet, we should keep in mind that the database provides the following information for each Lexical Unit:

(a) monolingual information: corpus-attested examples, semantic frame, cognitive motivation, semantic prosody, syntactic-semantic profile, collocation, marked usage

(b) bilingual information (derived from a contrastive analysis of monolingual corpus data): translation equivalents, translated examples, SL-TL contextual patterns, notes on the degree of correspondence between prima facie equivalents.

Focusing attention on the synthesis stage of producing a particular dictionary, we proceed to provide a preliminary description of the proposed dictionary and to delimit the designing task. On the basis of Atkins & Rundell’s (2008: 24-25) list of criteria for classifying dictionaries, Figure 2.1 summarizes the main properties of the dictionary in terms of languages, coverage, medium, organization, users, and uses. The scope of the present study is limited to the Greek-English encoding function; yet, the electronic medium offers the possibility of incorporating many functions (e.g. monolingual, multilingual, encoding, decoding) into a single user interface. Therefore, the proposed dictionary design could lead to a lexicographic product for a specific market or be integrated into a multi-functional reference work.

Figure 2.1. Properties of the proposed dictionary [1]

In deciding how information available in the pre-lexicographic database should be presented in the dictionary, and whether additional material should be specially compiled, we take account of relevant developments in e-dictionary design. User-friendliness is closely related to planning issues such as modes of data access, new components (e.g. thesaurus- or corpus-like material, language awareness articles), the display layout (e.g. stratified interface levels), and navigation (e.g. menus, pull-down and pop-up windows, built-in tutorials on how to use the dictionary) (de Schryver 2003; Sobkowiak 2007; Atkins & Rundell 2008: 238-246; Pastor & Alcina 2010).

To give an example from the highly competitive (and hence innovation-seeking) market of EFL lexicography, the e-MEDAL2 (2007) takes advantage of the flexibility of the electronic medium (CD-ROM) to offer a wide array of search options (e.g. search for words by the way they sound, or for words with extra features, use filters like grammar, style, frequency, etc., search within the example sentences in the A-Z dictionary text, or within the “Study Pages” section). It also masks extra or advanced material (e.g. thesaurus, inflections, collocations, usage notes, metaphors, word stories, weblinks, sound effects, animations, exercises), reducing thus user disorientation.

In this light, Figure 2.2 sketches out the improvements and innovations that the proposed dictionary design can effect on the scene of Greek-English lexicography. Benefits are derived from utilizing a pre-lexicographic bilingual database, a predetermined user profile, and the electronic medium potential. The underlying database allows for an improved treatment of lexical meaning and use, and for a principled division of senses. The user profile determines the orientation of the dictionary reflected particularly in the metalanguage and in the amount of information on the TL. The electronic medium enhances accessibility, and accommodates new entry components, supplementary resources, and customization options.

Figure 2.2. The dictionary design at a glance

3. Establishing a user profile for the proposed dictionary

In order to achieve the right fit between dictionary type and user needs, we need to establish a clear picture of the prospective users. This section presents the user profile of the proposed dictionary, considers helpful findings on target user groups, and illustrates how user-orientation is integrated into the proposed dictionary design.

On the basis of Atkins & Rundell (2008: 28-29), the following user profile has been drawn up for the proposed e-dictionary. As Figure 3.1 shows, three types of information are specified: types of user, types of use, and users’ pre-existing (linguistic and reference) skills. Undertaking user research (e.g. questioning users, or setting up experiments to monitor dictionary use) is beyond the scope of this work; available findings on dictionary use (for production purposes) by learners and translators help us gain a clearer view of target users.

Figure 3.1. User profile for the proposed dictionary [2] [3] [4] [5] [6]

Relevant in this respect is Laufer’s (1995) proposal for creating a special L1-L2-L2 dictionary for production, and Laufer & Levitzky-Aviad’s (2006) consequent empirical investigation into the usefulness of such a dictionary –called “bilingual dictionary plus” (BD+)– for foreign language writing. The study on the effectiveness of a BD+ Hebrew-English-English experimental dictionary (in paper and computerized version) has demonstrated the superiority of the BD+ over an English-English-Hebrew dictionary and a Hebrew-English one. It has also revealed the “translation + definition + example” lookup pattern as the most prevalent one, and the electronic version as the most preferred one.

Besides writing in a foreign language, user research has also been interested in dictionary use for the purpose of L1-L2 translation. According to Varantola’s (2003) small-scale study, translators habitually turn to both bilingual and monolingual sources looking for contextual, pragmatic and encyclopaedic information. Pointing in the same direction, Atkins & Varantola’s (2008: 362-366) larger-scale study shows that the examples section of bilingual entries and the information available about L2 collocation are not sufficient to lead users to a successful translation of problematic items like context-dependent general words and multi-word expressions (MWEs). It is thus suggested that “the combination of a comprehensive bilingual dictionary […] and a selected monolingual L2 corpus, in which the examples are broadly differentiated according to word senses, would offer advanced language practitioners a powerful translating tool” (ibid.: 366); such a tool is envisaged for electronic access only (ibid.: 371).

The present work draws on the above insights gained from research on dictionary use in order to develop an e-dictionary design tailored for Greek-English (encoding) users. To demonstrate the role of user-orientation in the design, Figure 3.2 shows the special benefits that the main user groups addressed can obtain from the proposed e-dictionary.

Figure 3.2. User-orientation in the dictionary design

4. Describing the user interface of the proposed dictionary

Το describe the user interface of the new e-dictionary, we graphically depict its main screen layout in Figure 4.1. The menu bar consists of navigation arrows, operation icons, and buttons directing to options in drop-down menus; the options are indicated by dotted arrows and tables.

Under the menu bar there is the search box, inside which a particular word can be typed. By pressing the “Search” button («Αναζήτηση») or the Enter key on the keyboard, the search for the particular word is performed; the dictionary has an auto-complete feature, i.e. “when the user starts introducing a word in the dictionary, a drop-down list appears suggesting entries included in the dictionary that start with the letters that the user is introducing” (Pastor & Alcina 2010: 321). The headword(s) found is/are displayed underneath in the result panel («Αποτελέσματα»); by clicking on a headword in the result panel, the dictionary entry is displayed in the main panel.

By way of explanation, the electronic functions of the proposed dictionary are described in the following seven units, which present the elements of the menu bar from the left to the right.

Figure 4.1. Dictionary menu bar buttons and options

I) arrow icons on the left of the menu bar:

By pointing the cursor on the icons, users obtain an instant explanation of their respective function, i.e. «εμπρός», «ιστορικό», «πίσω», as Figure 4.1 shows. The History button enables users to review a list of the recent dictionary entries they have visited, while the Back and Forward buttons allow them to step back and forth through the entries on the list. These buttons facilitate navigation and allow for an arrowed search route.

II) search options («επιλογές αναζήτησης»)

The Search Options button displays a drop-down menu, from which options can be selected to carry out an advanced search, to search within the Examples Bank section, and to browse an alphabetical or a thematic index.

(a) Advanced Search («Σύνθετη Αναζήτηση»):

By selecting this option from the menu, the Advanced Search window is displayed, which would look like Figure 4.2.

In the search box users can enter the words they wish to search for. Users can type lemmas or word forms, in M. Greek or English, and they can search for phrases by enclosing words in double quotes (“”). It is possible to use regular expressions and Boolean operators in queries, in order to replace one or more characters of a word and to search by a combination of words.^[7] In the Advanced Search window, the resource that can be queried is not limited to the headword list; rather, it includes the full text of the dictionary (menu boxes, sense indicators, equivalents, translated example sentences), as well as the separate entry component Notes on Equivalence (see section 5). Users can further filter their search by various criteria, e.g. part of speech, non-literal use, style, region, and scientific field. Search criteria (in the form of preordained codes) are selected from the drop-down lists provided; in the encoding function of the dictionary, these criteria refer to the TL (i.e. English).

By pressing the «Αναζήτηση» button, the list of entries (or entry sections) that match the search criteria appears on the left-hand side of the results panel.^[8] The entry (or entry section) selected is displayed on the right-hand side of this panel, and the items queried are highlighted. Icons are also provided for guiding users to copy, save and print the results.

Figure 4.2. The Advanced Search window

For example, by typing the combination πετάω + ευκαιρία and checking the boxes «λήμμα» and «αναζήτηση και στο σώμα των λημμάτων», the menu box of the πετάω entry appears (see section 6), and two hyperlinks are highlighted (i.e. «για χαμένη ευκαιρία» and «σπαταλάω λεφτά/ ευκαιρία») leading users to the relevant sense divisions. To illustrate the function of filters, we may use the following example. If we typed the phrase “πετάω έξω” in the search box, checked the boxes «λήμμα» and «αναζήτηση και στο σώμα των λημμάτων», and asked for a TL item of informal style, the dictionary would show the relevant sense division of the πετάω entry (i.e. «απαλάσσομαι από σκουπίδια/ κάποιον ανεπιθύμητο» in Figure 6.1) and highlight the translation options kick out and boot out.

(b) Examples Bank Search («Αναζήτηση στην Τράπεζα Παραδειγμάτων»):

The Examples Bank is a dictionary-external TL resource beneficial to advanced users. It consists of corpus-attested English example sentences drawn from the underlying monolingual database (see chapter 2); it is not a bank of the translated Greek example sentences. The Examples Bank search button gives access to a new window (see Figure 4.3) with a simple corpus query system for users to search for words within the underlying monolingual (TL) database. Wild cards and Boolean operators can be used in the search, and results can be sorted to the node (requested word), its right or left context. The default display of results is in the form of sentences, but the option of displaying concordances is also available.^[9]

The Examples Bank is also integrated into the bilingual dictionary for directly providing further information about TL use in context. Each dictionary entry includes hyperlinks to examples in English that show each translation equivalent (in its specific sense) in context (see section 5, VI).

Figure 4.3. The Examples Bank search window

(c) Alphabetical Index («Αλφαβητικό Ευρετήριο»):

By pressing this button, users can browse an Alphabetical Index of the Greek headwords in a new window. Clicking on a headword will display the corresponding Greek-English entry in the main view mode.

(d) Frame Index («Θεματικό Ευρετήριο»):

By pressing this button, users can browse a Frame Index in a new window which allows queries guided by concepts rather than lexical items.^[10] The list of semantic frames looks like the one provided by FrameNet (https://framenet.icsi.berkeley.edu/fndrupal); in this case, though, M. Greek is the presentation language. Clicking on a semantic frame would display a two-column table with Greek and English lexical items (not necessarily direct equivalents to each other) that evoke this frame. Then, a multi-path interface could allow for the display (via hyperlinking) of the corresponding Greek-English and English-Greek entries for contrast purposes. The full design and implementation of this feature is outside the scope of the present study.^[11] Yet, it should be specified that this resource will draw its data from the frame-semantic analysis of the underlying database (see chapter 2). The Frame Index window could be structured as shown in the illustrative example of Figure 4.4.

Figure 4.4. The Frame Index window

The Frame Index supplementary resource should be carefully designed to provide access to comparative subject-oriented notes. For instance, Figure 4.4 shows that it would be useful to incorporate three hyperlinks in the self-motion frame. Taking advantage of multimedia technology, the «Απεικόνιση» window would list animations that represent the actions indicated by the motion verbs in a non-static manner.^[12] The second hyperlink would open a grammar note («Τρόπος Κίνησης & Γλωσσική Έκφραση») on the systematic mismatch of semantic distribution between the two languages, thus explaining why the lexical set of English manner-of-motion verbs is larger. The third hyperlink would open a grammar note («Προθέσεις Κίνησης») on the use of English prepositions and particles. The last two devices essentially link the Frame resource to the Study Pages resource.

III) study pages («μελέτη»)

The second button of the menu bar opens the Study Pages resource in a new window. This resource displays a series of articles (in M. Greek) which aim to (a) enhance users’ language awareness, (b) improve their writing skills, and (c) inform them about the bilingual dictionary function and use. The idea of incorporating such a resource into the e-dictionary is based on the e-MEDAL2. Although the actual compilation of the Study Pages lies outside the scope of the present study, we should specify their contents and access modes.

The language awareness section would contain articles on metaphor/ metonymy, polysemy, co-occurrence patterns, and pragmatics. These articles would look like the Study Pages in the e-MEDAL2 except that they would be in M. Greek. The second section, which would be designed to help users become more effective writers, would include articles on communicative functions and grammar. This section concerns English (L2), uses M. Greek (L1) as the presentation language, and draws attention to differences between the two language systems; it would resemble the English Grammar Guide included at the end of the CGED, except that it would contain more interlingual contrasts so that users could avoid the indiscriminate use of word-to-word translation in encoding. Lastly, information would be provided about the notion of “equivalence”, about the function of bilingual dictionaries in general, and about the information categories in the particular e-dictionary; these articles on dictionary use would function as an introduction in a print dictionary. On the basis of this description, we could recommend the following names for the three sections of the Study Pages resource:

- «Γλωσσική Επίγνωση» (containing the articles: «Μεταφορά & Μετωνυμία», «Πολυσημία», «Φρασεολογία», «Πραγματολογία»)

- «Παραγωγή Γραπτού Λόγου» (containing the articles: «Θέματα Επικοινωνιακής Λειτουργίας», «Θέματα Γραμματικής»)

- «Το Δίγλωσσο Λεξικό» (containing the articles: «Μετάφραση και ‘Ισοδυναμία’», «Πληροφορίες στο Λεξικό»).

To facilitate navigation, the Study Pages window should be equipped with buttons to jump between the articles, and to go back to the contents of this resource. Apart from this ‘navigation’ mode for accessing the articles, a ‘search for’ mode should also be available. In addition, the Study Pages need to be carefully integrated into the bilingual dictionary; although they essentially constitute a dictionary-external resource, they should not merely be placed alongside the dictionary. A step towards integration can be made by consistently hyperlinking Study Pages to relevant metalinguistic labels in the entries (see section 5, III), as well as to members of lexical sets. For example, consider the following potential points of access from the dictionary entries to the Study Pages:

-  non-literal use labels (e.g. μεταφορικη χρηση, μετωνυμικη χρηση) → Study Pages on metaphor/ metonymy, and polysemy

-  usage patterns labels (e.g. εκφραση, περιφραστικό ρήμα) → Study Page on phraseology

-  style labels (e.g. ανεπισημο υφοσ, επισημο υφοσ) → Study Page on pragmatics

- motion verbs (in their literal sense) → Study Pages on the different lexicalization patterns of motion in M. Greek and English, and on the use of English particles and prepositions (grammar notes: «Τρόπος Κίνησης & Γλωσσική Έκφραση» and «Προθέσεις Κίνησης»).^[13]

IV) exercises («ασκήσεις»)

The Exercises button opens the contents of the Exercises section. Again, the actual compilation of this resource lies outside the scope of the present study. However, it should be explained that Exercises are expected to aim at the following: 

- to provide practice for the information given in the Study Pages (especially, raising awareness of polysemy and metaphor)

- to train users to make good use of the bilingual dictionary for production tasks; interactive translation tasks would be developed to give users practice in choosing the appropriate English equivalents and finding how to use them in context

- to provide practice for the ‘mediation’ writing test of the exam for the Greek State Certificate of Language Proficiency at B1, B2, C1 levels;^[14] the mediation task lends itself to the encoding use of bilingual dictionaries because it tests the ability to produce written language based on input in M. Greek.

All exercises should be marked for their level of difficulty (intermediate/ advanced), so that users can make the choice they wish. For closed tasks, answers would be checked interactively, while for open tasks (e.g. text translation, mediation) sample answers would be available. The Exercises window should be equipped with navigation buttons and allow for copy, save, and print operations, so that exercises could be easily turned into classroom worksheets. 

V) settings («ρυθμίσεις»)

The Settings button accesses a drop-down menu (shown in Figure 4.1 above) that provides users with options for adapting text size and information amount to their preferences. Users can uniformly:

- increase, decrease, or normalize (again) the font size

- select the display of extra information sections (useful TL patterns of co-occurrence, notes on equivalence, examples bank), i.e. advanced material which is masked by default (see section 5)

- deselect the display of menus and labels (see section 5, I & III); these features are displayed by default because they facilitate navigation and choice of the appropriate translation equivalents

- restore the default settings

The Settings options aim at achieving a degree of customization, interactivity and adaptability, as well as stratified interface levels.

VI) help («βοήθεια»)

The Help button provides access to a demo tour of the e-dictionary and a user guide in M. Greek (see Figure 4.1). This user guide can be used in combination with the Study Pages on dictionary use («Το Δίγλωσσο Λεξικό») to compile the dictionary documentation for the user, i.e. the printed materials that should accompany the software disk.

VII) icons on the right of the menu bar:

By pointing the cursor on the icons, users obtain an instant explanation of their respective function, i.e. «αντιγραφή», «αποθήκευση», «εκτύπωση», «προσθήκη σημειώσεων» (see Figure 4.1). The Copy button, the Save button, the Print button and the Add Note button facilitate customization. A dictionary entry or a selected text can be copied, saved and printed. The Add Note button opens a new window and offers users the possibility to add their own notes to dictionary entries, edit, save, or delete them.

5. Describing the entry structure of the proposed dictionary

This section complements the picture of the user interface by presenting the structure of the dictionary entries. Figure 5.1 uses English metalanguage and colour to show the information categories to be displayed in the main view mode of the production-oriented e-dictionary.

More precisely, six main entry components are provided: (i) menu, (ii) SL sense indicators, (iii) TL translation equivalents, (iv) translated example sentences with useful TL patterns of co-occurrence, (v) notes on equivalence, (vi) TL corpus. A description of these components follows, and the corresponding Greek labels, which would appear in the user interface, are indicated. It should also be noted that in Figure 5.1 colours are functional rather than cosmetic; they serve as attention getters which highlight key words and discriminate information categories. Throughout the dictionary text, Greek headwords appear in bold black italics and English translation equivalents in bold red italics. The example sentences, which show the SL and the TL in context, are placed in a blue box. In grey background appear the three extra boxes, i.e. the Menu, the Notes on Equivalence, and the Examples Bank.

Figure 5.1. Information categories in the e-dictionary entries

I) menu («μενού»)

Long dictionary entries (for polysemous Greek headwords) have a Menu at the top to make it easier and faster for users to spot the Greek sense or MWE they wish to express in English. The Menu, which is a very familiar concept to the average computer user, is displayed in the users’ native language, i.e. M. Greek. To create a kind of “table of contents” for a dictionary entry, we gather sense indicators and MWEs in a single block at the top of the entry proper. Each sense indicator and MWE in the Menu functions as a hyperlink that directs to, and highlights, the right part of the bilingual entry. Menu boxes are displayed by default, since they contribute to a more user-friendly interface by facilitating findability and navigation; however, users have also the option to manually close them (via the Fold button [+]/ [-]), or to uniformly deselect them in the Settings dialogue box (see section 4, V).

Menu boxes have become rather well known from the electronic editions of English monolingual learners’ dictionaries; yet, in combining menu-guided sense access with target sense highlighting, we draw on Lew & Tokarek’s (2010) study on the effectiveness of entry menus in bilingual e-dictionaries. These authors explain that target sense highlighting is a new presentation device, which is afforded by modern advances in information technology, and offers significant benefits beyond the bare menu, both in speed and accuracy. In the present dictionary design, the device of highlighting is thought to be particularly beneficial for accessing MWEs. With a view to facilitating scanning, MWEs are recorded in alphabetical order in the Menu (under the heading «Στερεότυπες Εκφράσεις»), but they are placed within the related sense division in the entry text to indicate semantic motivation; highlighting thus helps users locate the target MWE without scrolling.

II) sl sense indicators

The dictionary senses appear numbered in the main panel. Headword sense indicators are systematically employed as a structuring device; they are in M. Greek and may be of three types:

- specifiers (….), i.e. a paraphrase or synonym of the SL sense

- collocators [….], representing whole semantic classes co-occurring with the SL headword in its specific sense

- domain labels, such as <οικονομία>; these should be sparsely used because they are considered as less user-friendly than the other sense indicators (Atkins & Rundell 2008: 215).

III) tl translation equivalents

Under the indicator of a SL sense appear the TL translation equivalent(s) together with a series of informative hyperlinks and labels. Users can click on the «προφορά» hyperlink to hear the English word spoken, while the «κλίση» hyperlink displays its inflections in a new window. The labels used specify the English equivalents; they appear in Greek metalanguage and can indicate the following:

- part of speech (e.g. ρήμα) or phrase type (περιφραστικό ρήμα, έκφραση)^[15]

- transitivity for verbs (μεταβατικό/ αμετάβατο) or countability for nouns (μετρήσιμο/ μη μετρήσιμο)

- non-literal use (μεταφορική χρήση, μετωνυμική χρήση)

- style (e.g. ανεπίσημο ύφος), region (βρετανικός τυποσ/ αμερικάνικος τύπος); marked style is also indicated for SL to reassure users, who should assume that the TL translations will match the SL unless indicated otherwise.

Pointing the cursor at these labels displays pop-up notes briefly explaining the messages carried in the metalanguage, while double-clicking on the έκφραση, περιφραστικό ρημα, μεταφορικη χρηση, μετωνυμικη χρηση, and ύφος labels displays the hyperlinked Study Pages (see section 4, III). The labels are displayed by default because they guide users to choose the appropriate translation equivalent; yet, if users prefer a simplified view mode they can switch them off in the Settings dialogue box (see section 4, V).

IV) sl example sentences + their tl translations + useful tl patterns of co-occurrence («παραδείγματα» + «μετάφραση» + «τυπικές συνεμφανίσεις»)

Information about use in context is presented under the translation equivalents, in a blue table, which by default displays Greek example sentences and their translations in English. This type of information is to be directly drawn from the underlying pre-lexicographic database (see section 2), and is organized in two columns (i.e. «Παραδείγματα» and «Μετάφραση»). The table is adaptable; more or less translated examples can be shown as the Fold button ([+]/ [-]) indicates.

An additional information category («Τυπικές Συνεμφανίσεις») can be displayed in a third column next to the basic table. This innovative column is specifically designed to list explicit pattern illustrations to be used as models in producing appropriate texts in English – an encoding type of information traditionally available only in monolingual learners’ dictionaries. Users have the option to open/ close the column pressing the Fold button ([+]/ [-]), or through the Settings dialogue box (see section 4, V).

The blue table attempts to make SL and TL combinatorial properties explicit. The design of this entry component, as well as of the next one (i.e. Notes on Equivalence), has been driven by (a) Sobkowiak’s (2007: 43) call for dictionary interfaces that highlight lexical patterns that are difficult or impossible to be seen in linear form, and (b) Yong & Peng’s (2007: 207) recommendation that “it is particularly important to reflect the habitual co-occurrence and remind dictionary users to avoid possible lexical mismatches and traps in the active bilingual dictionary”.

V) notes on equivalence («σχόλια ισοδυναμίας»)

In the right-hand panel next to each numbered dictionary sense there is a minimized “Notes on Equivalence” box. If we press the Fold button ([+]), local notes are displayed concerning the degree of correspondence between prima facie equivalents. Remarks are made on style, metaphoricity, and translation strategies; cross-linguistic differences and regular patterns, which are impossible to be seen in linear form, are foregrounded. This type of information is available in the underlying pre-lexicographic database (see section 2); yet, it needs to be presented in a comprehensible and user-friendly manner with its metalanguage in M. Greek.

In this way, the proposed bilingual e-dictionary would provide SL users with the “bridge” to the TL without nurturing the common misconception of exact equivalence, or swamping users with complex codes for indicating lack of equivalence. Depending on their background/ skills and the nature of the production task, users can uniformly select or deselect the Notes on Equivalence boxes through the Settings dialogue box (see section 4, V).

VI) examples bank («τράπεζα παραδειγμάτων»)

In the right-hand panel next to each numbered dictionary sense and below the Notes on Equivalence box there is a minimized “Examples Bank” box. By pressing the Fold button ([+]), users are provided with a list of hyperlinks to corpus attested sentences, which show the suggested English translation equivalents (in their specific sense) in use.^[16] Advanced users can thus obtain additional contextual information about the TL by accessing a complementary corpus. In other words, the proposed encoding e-dictionary gives access to a back-up TL resource, which draws on the underlying database compiled from raw corpus data (see section 2). Users can also carry out independent searches in the Examples Bank through the «Αναζήτηση στην Τράπεζα Παραδειγμάτων» button in the Search Options («Επιλογές Αναζήτησης») menu (see section 4, IIb).

6. Sample from an entry

The lemma used for illustrating the bilingual entry design is πετάω ’fly’. Figure 6.1 presents a Menu of the relevant SL sense indicators and MWEs in hyperlink form. Figure 6.2 shows the content of the third hyperlink, i.e. a sample section from the entry which exemplifies both matches and mismatches between the apparently direct equivalents πετάω - fly. Two points need to be specially noted, i.e. the arrangement of MWEs, and the entry structure.

A glance at the Menu would reveal what MWEs have been entered in the πετάω entry; having a close look at the entry, users could then realize which MWEs are motivated by the self-motion and which ones by the cause-motion core meaning of the verb. For example, πετάει η ομάδα appears in the self-motion network of πετάω under the section that concerns “increase in speed/ value/ performance” (3γ). To provide a clear picture of the different collocate types in the sense divisions, we use a tiered entry structure (e.g. 3α. [vehicle, human], 3β. [amount, value], 3γ. [process, team]).

πετάω/ πετώ^[17]

Figure 6.1. The Menu box in the entry for πετάω

Figure 6.2. Sample section from the entry for πετάω

7. Conclusion

In designing a novel bilingual e-dictionary that meets users’ encoding needs, we have focused on the role of the dictionary in the reference process. The proposed features and functions (e.g. stratified interface levels, new entry components, integrated supplementary resources, customization options, explicit metalanguage in L1, large amount of information on L2) can potentially be of great benefit to the main user groups addressed –i.e. EFL learners (school students, literate adults) and language professionals (EFL teachers, coursebook writers, translators)– in various encoding tasks, such as translating Greek text into English, expressing themselves in English, learning/ teaching EFL in the Greek context, and preparing for the Greek State Certificate of Language Proficiency exams.

Nevertheless, problems in dictionary consultation arise not only due to defects of dictionaries, but also due to users’ poor reference skills (Hartmann 2001: 89). The implication is that users need to be trained to take full advantage of the wealth of information and facilities provided. Various aids, such as user guides, dictionary workbooks and formal instruction, can prove useful in this direction.

About the author(s)

Thomai Dalpanagioti has graduated from the English Department of Aristotle University of Thessaloniki, Greece, and holds an MA and a PhD degree in Lexicography from the National and Kapodistrian University of Athens, Greece. She is currently an EFL teacher in Greek secondary education and a researcher in the Centre for the Greek Language (Hellenic Ministry of Education). Her main research interests concern translation equivalence, polysemy and phraseology in corpus-based bilingual lexicography.

Email: [please login or register to view author's email address]

•  Previous article
• Next article