One of the main aims of this research is to propose an entry structure of the future dictionary which would treat abbreviations in a more appropriate way. By appropriateness, for one thing, we mean the way(s) an entry is organized. Majority of abbreviation dictionaries are unidirectional and explanatory and do not deal with other information a user would find interesting and useful, such as information on diachronic, diatopic, diamesic, diastratic and diaphasic variations. Here it is necessary to say that not all variations are of the utmost importance for a dictionary of abbreviations, but some, like diachronic, diamesic and probably diatopic variations, would prove to be valuable in making its scope more comprehensive and informative.
We shall now briefly consider the main objectives in compiling the dictionary, i.e. its main characteristics. Due to the fact that the spelling of the headwords would, for the most part, consist of capital letters, we believe that the alphabetical arrangement of lemmata would be the most suitable form of presentation for this dictionary. The term bilingual implies a dictionary that would explain abbreviations in both English and Croatian, which means it would be bidirectional: firstly explaining an English abbreviation in English and secondly, providing a translation of the expansion in Croatian, as well as the translation of information on the diamesic level. The feature of being specialized is understood within the frame of a scientific and technical dictionary which would firstly serve as a lexicographic tool for professionals in the field, synchronically representing the specialized lexicon as it exists. Secondly, it would also be aimed at both wide, general, non-specialized and specialized audience. The singularity of the subject-field indicates the purpose of the dictionary to be informative and encyclopeadic in content for that particular scientific field. The specialized corpus of the dictionary identifies the domain- specific collection of abbreviations in linguistics from specialized publications in the field. As for the final explanatory feature, the entries would present the following data: spelling, pronunciation, typology of abbreviations, full form of source phrases (expansions), orthographic and morphological characteristics of both abbreviations and their expansions, and the origin of headwords.
2.1. Two-sidedness of the lemma
Contrary to previous lexicographic practice of dealing with the micro-structure in a dictionary of abbreviations, we feel that headwords in such a dictionary should be considered differently from those in a general-purpose dictionary, primarily because of the specific formational processes inherent to abbreviations. In other words, their headword form should be seen as two-sided, which means that if we are to understand not only the meaning of an abbreviated form, but its final orthographic form as well, we have to take into account both sides – the abbreviated form and its source phrase or expansion (e.g. ACL – 'Association for Computational Linguistics'). The left side of the lemma represents its final abbreviated form while the right side its expansion. The right side of the lemma is not merely its meaning, but is also its morphemic structure (although its formation is sometimes considered to belong to non-morphemic or non-morphematic word-formation processes, cf. Fandrych
2004). Having taken into consideration its two-sidedness and various characteristics of forms found in the right side of the lemma, we shall see that this would necessarily lead to a unique designation of their distinctivness. For the purpose of their differentiation, a system of exclusive classification and subclassification of abbreviations proposed in earlier research (cf. Fabijanić, Malenica 2013) is improved in this research.
2.2. Narrower and broader sense of the lemmata
Abbreviations in the linguistic terminology will be classified according to two main criteria – narrower and broader sense. The narrower sense of their creation refers to those formed by using initial letters of each element in the expansion (mainly alphabetisms), and pronounced either by individual names of letters or as a word. The broader sense implies the ways and processes of formation, more or less different from the orthographic norms (mainly hybrid forms, acronyms, blends and clippings featuring some orthographic changes), in consequence of which, one or more initials are used for various smaller elements of the expansion (smaller than words, yet bigger than initials). Due to this, initials for graphemes, compounds, affixes, grammatical and lexical words found in the final form of an abbreviation, as well as different orthographic changes, such as ellipsis, conversion, metathesis, addition, etc., will be analysed and (sub-)classified. By considering their different expansion combinations and comparing those different combinations with the final abbreviated forms, we will be able to understand the real nature of their (non-) morphemic structure and the possibility of differentiating the initialisms according to various criteria.
2.3. The multi-level approach
A tentative solution for the lexicographic presentation of abbreviations is based on Fandrych's (2008a) theoretical framework, i.e. the multi-level approach for the analysis of submorphemic word-formation processes. The multi-level approach is comprised of three main aspects: 1) Structure and Modes of Production, i.e. the structural aspects and word-formation potential, word class, medium and origin; 2)
Cognitive Aspects, i.e. semantic, semiotic and motivational aspects, lexicalisation
and institutionalisation, and 3) Functional Aspects, i.e. stylistic and sociolinguistic aspects, pragmatic and text-linguistic aspects. We believe that the application of this interdisciplinary approach will give a fuller and more transparent picture of various orthographic, morphological, semantic, stylistic and functional processes involved in the production and uses of abbreviations.
As for the initial step in compiling this specialized dictionary, the first aspect of the multi-level approach, i.e. the structure and modes of production, will be supplemented by our novel aspects within the structure and modes of production, i.e. on both orthographic and morphological level. It has been observed that "[t]here is evidence for a remarkable word-formation potential, in the form of multiple formations…[and] there are also a number of non-morphematic word-formation processes which are multiple formations." (Fandrych 2008a: 74-75; emphasis in original). The word-formation potential in this work is understood exactly as the
reapplication of the bi-aspectual criterion used in the creation of abbreviations, i.e. the criteria of narrower and broader sense. If one of the criteria is being used more frequently than the other, most frequently narrower over broader sense, it obviously represents the greater word-formation potential of abbreviations.
2.4. The entry and its elements
Firstly, an entry would consist of a headword and all its possible variants. Together with the data about its origin and the medium the abbreviation was noted in, the information on the word class (where applicable) will be also provided. Basic form would be succeeded by the information on pronunciation and the type of word-formation. A whole set of novel descriptors will be used when dealing with the type of word-formation (i.e. the modes of production). The descriptors will inform the user about the originality and regularity of the final word-forms, i.e. whether they were formed according to the orthographic norms of abbreviations or with some evident exceptions to the norms.
For the purpose of their differentiation we proposed a system of exclusive classification and subclassification of abbreviations (Fabijanić, Malenica 2013). Miscellaneous realisations of abbreviations are generally diversified into two main groups: those considered as realised in narrower sense and those in broader one. Abbreviations in narrower sense are exclusively explained with an LLL descriptor for initials used in their formation (a characteristic three-letter descriptor was chosen due to the most frequent number of graphemes found in most initialisms). Abbreviations in broader sense are represented with a whole set of additional different letters or initials (written either in capital or small letters) added to a three-letter descriptor: e.g. l for initials made of small letters, P refers to initial affixes (mostly prefixes), N stands for a numeral, S refers to a syllable, and W stands for a word used in a hybrid form of an abbreviation. Orthographic changes, evident by comparing initialisms (alphabetisms or acronyms) with their expansions, are explained by other descriptors: E for ellipsis, C – conversion, M – metathesis, and A for addition of a word or a diacritic sign not normally found in expansions. Comprehension and consequently classification of abbreviations depends on the degree of their (non-)coordination with the common orthographic norms.
The following examples of different abbreviation forms (an alphabetism, an acronym, a clipping, a blend and a hybrid form) present the micro-structure of a future dictionary entry, i.e. the structure in which all elements of the first aspect (cf. § 2.3. The multi-level approach) were taken into consideration.
ADS [] T: alph.E: American Dialect Society/Američko dijalektalno društvoM:
written/ pisani, spoken/govorniD: LLLO: ADS was founded in 1889 with the intention
of creating a dictionary of American dialects. (ELL)/Američko dijalektalno društvo osnovano 1889. s namjerom stvaranja rječnika američkih narječja.
ACE [] T: acr.E: 1) Automatic Content Extraction/Automatsko ekstrahiranje
sadržaja, 2) Australian Corpus of English/Korpus australskoga engleskogM:
written/pisani, spoken/govorniD: 1) LLL, 2) LLL E = prep.; O: 1) The ACE program is a
successor to MUC that has been running since a pilot study in 1999. (ELL)/ACE program provodi se još od pilot-projekta iz 1999., a naslijedio je MUC., 2) The corpus of
Australian English compiled at Macquarie University using texts published in 1986. (HEL) / Korpus australskoga engleskog sačinjen na Macquarie sveučilištu iz tekstova objavljenih 1986.
AUX, Aux, aux [-] T: clip.E: Auxiliary/pomoćniM: written/pisaniD: LllO: Lat.
auxiliaris - 'giving aid' (RDLL) /lat. pomoćni - 'koji pomaže'.
AFRILEX [] T: blendE: AFRIcan association for LEXicography/Afričko
leksikografsko udruženjeM: written/pisani, spoken/govorniD: LSF E= noun, prep.O:
AFRILEX was founded in 1995 and strives to promote all aspects of lexicography on the
African continent. (ELL) / Organizacija osnovana 1995. u svrhu promocije svih aspekata leksikografije na afričkom kontinentu.
ALGOL [] T: hybr. (syll+s.abb.)E: ALGOrithmic Language/algoritamski jezikM:
written/ pisani, spoken/govorniD: SSLO: Programming computer language appeared
in 1958. (RDLL) / Programski računalni jezik nastao 1958. godine.
Each entry consists of seven structural elements, five of which are represented by abbreviations (simple abbreviations): T, E, M, D, and O. The entry starts with a headword and its variant forms, all written with either small or capital bold letters, and is followed by its pronunciation with the use of IPA symbols. The first abbreviation (T) stands for the type of abbreviation explained and described, and relies on our subclassification of abbreviated formations (cf. § 2.5., 3.). E stands for the expansion elements that abbreviations were formed from. M represents the medium a headword was used in – written, spoken or both media. D represents a previously determined set of descriptors used in the classification of abbreviations according to the expansion elements compared with the resultant final form. O, the last user abbreviation, refers to the origin of the headword, i.e. to the information about the date of its emergence, the author/-s who devised it or the language it was borrowed or adapted from. The information given in the elements about the medium and the origin is also provided in Croatian.
2.5. Previous research on orthographic features of abbreviations, morphological features of expansions, and the correlation with the recent study
The previous research of abbreviations in English medical terminology (Ibid.) has justified the use of the mentioned classification. Most of the alphabetisms were formed according to the criterion in narrower sense and formed of initials written with capital letters which were taken from every single element found in the expansion. Acronyms were fewer and the ratio of those formed in narrower and broader sense was not as regular as for alphabetisms, i.e. in favour of those formed in narrower sense. The ratio for alphabetisms was 124 : 53 (i.e. they are 2,3 times more frequent) and for acronyms – 10 : 11, which means that they are almost equally represented. As far as the ratio of hybrid forms is concerned, the irregularity is even more evident – 2 : 9 in favour of broader-sense formations (i.e. those formed by narrower-sense criterion are 0,2 times less frequent).
The research of abbreviations in linguistics has presented us the following statistics: the ratio for alphabetisms is 159 : 109, in favour of narrower sense (they are 1,4 times more frequent than those formed in broader sense), for acronyms – 23 : 31 (0,7 times lesser frequency of those formed in narrower sense), for simple abbreviations the ratio is 19 : 1 for narrower-sense formations, and for clippings – 15 : 6 (i.e. they are 2,5 more frequent than the broader-sense formations). Blending has only been attested through the examples of broader-sense formations. A similar ratio for hybrid formations, in which broader-sense abbreviations are more frequent than narrower-sense ones, is also evident in this research.
3. The corpus of abbreviations in linguistics and the methods of analysis