Through the lexical labyrinth
Using polysemy and a large general corpus to introduce general technical vocabulary
By Andy Cresswell (University of Bologna, Italy)
Abstract & Keywords
Whatever the theory, in practice many translators and interpreters starting their careers will make a living from technical translation towards their L2. This implies the need for the acquisition of technical vocabulary in real L2 contexts rather than from bilingual dictionaries. To this end, learners explored polysemes in different technical contexts using a large general corpus, working along learner-centred, co-operative lines. Results showed the method gave learners experience of different technical fields, of “general” technical terms, of term-formation, and of intensive vocabulary gathering, and was effective in highlighting corpus use as an alternative to the dictionary.
Indipendentemente dalla teoria, molti traduttori e interpreti iniziano la propria carriera gudagnandosi da vivere con traduzioni tecniche verso la propria seconda lingua. ciò implica il bisogno di acquisire il vocabolario tecnico in contesti reali nella seconda lingua piuttosto che in vocabolari bilingui. A tale scopo, gli apprendenti hanno esplorato le polisemie in diversi contesti tecnici usando un ampio corpus generale, lavorando in modalità cooperativa e incentrata sul discente. I risultati dell’esperimento hanno mostrato come tale metodo abbia reso gli apprendenti familiari con diversi campi tecnici, con termini tecnici “generali”, con la formazione di termini e con l’apprendimento di an ampio vocabolario. Si è inoltre dimostrata l’efficacia di usare un corpus in alternativa all’uso del dizionario.
Keywords: corpus-based translation studies, translation into l2, corpus linguistics, translator training
©inTRAlinea & Andy Cresswell (2002).
"Through the lexical labyrinth"
inTRAlinea Special Issue: CULT2K
Edited by: Silvia Bernardini & Federico Zanettin
This article can be freely reproduced under Creative Commons License.
Permanent URL: http://www.intralinea.org/specials/article/1682
This study originated in the attempt to meet translating and interpreting students’ need for technical vocabulary input while providing a practical introduction to the British National Corpus, through projects in which groups of students followed polysemic words around it. The results reported derive not only from data in the form of vocabulary reference booklets produced by the students, but also from the learners’ views and from my own participant observations.
My role in helping to teach translation and interpretation is not to teach translation as such, but to teach language to support translators and interpreters. These learners are preparing to translate towards their L2 as well as towards their L1, a practice that, though it is sometimes regarded as controversial, reflects professional reality - as Weatherby (1998) put it, “in practice, more than half of many professional translators’ work is done into their L2”. In learning to translate or interpret towards L2 a fundamental premise is clearly that language intake should not be restricted to items found in predominantly bilingual dictionaries while translating, the likely consequence of such narrowed focus being a repertoire of target language exponents skewed towards forms cognate with L1; which, axiomatically, do not reliably constitute the most natural choice of synonym. Or, to put it in other ways, source-language focused acquisition results in “school translation” rather than professional translation (Gile 1995: 22), or what Duff (1981) calls “the third language”, a form of unnatural and inappropriate expression which falls between the source and target languages and is to be found only in translations.
Malmkjaer (1988: 3) sees the frequent absence of language input on translation courses as the result of a temporary preference for cultural and practical aspects, arguing that the pendulum should now swing back; Wilss (1996: 206) supports the inclusion of language learning on translation courses from the acquisition standpoint:
The inclusion of first-language and second-language learning in a translation teaching course may seem to be marginal or even superfluous, but in fact it is not, because the bulk of the students consist of compound bilinguals who have learned their native tongue primarily in a home setting and their second and third language in a school setting. As a rule (at least in Germany), their native tongue and second language linguistic and cultural knowledge is so weak that a good deal of native-tongue and second-language teaching and learning is necessary, before translation courses in the narrower sense of the word can be initiated.
This must be particularly so with regard to LSP, which, as Wilss (1996: 206) points out, dominates translation activities all over the world . Furthermore, much of this work, at least in Italy, is given to starting translators, who thus need to be expert in a wide range of fields, many of them technical, each one of which has its associated specific vocabulary, so that what is learned in one translation task is often not transferable to the next, and translation becomes a mapless labyrinth learner translators and interpreters are obliged to re-enter perpetually to find their way through anew. Register analysis, using corpora of narrowly specific fields, can provide some guidance in allowing learners to develop skills of textual pattern analysis and lead them to highly specific vocabulary of restricted use (Gavioli & Zanettin 2000), which is useful if translators work repeatedly in the same narrow field; but it cannot provide learners with a general lexical system to help them find their way through the maze constituted by constantly new technical contexts.
1. General Words and Polysemy
In LSP, however, several such systems have been proposed, including Martin’s academic vocabulary (1976) and Nation’s university word-list, the items on which, he implies (1990: 14-16), do not just occur in university texts. However, these systems refer essentially to educational settings; on the other hand, Dudley-Evans and St.John’s (1998: 83) revision of Baker’s 1988 categorisation of semi-technical vocabulary seems a more practical proposition in allowing itself to be adapted more readily to texts that were not written by academics or school teachers, like most of the texts that young translators have to translate. The scheme consists of (i) “general vocabulary that has a higher frequency in a specific field”, and (ii), “general English words that have a specific meaning in certain disciplines”.
The term “general word” is open to interpretation; more precisely, what appears to characterise these words is that they are polysemic. They are not homonyms - Lyons (1995: 58-9) distinguishes between homonymy, the coincidental resemblance of two distinct lexemes, and “polysemy”, or multiple meanings, in which the meanings are linked by a process of metaphorical extension; and this is what we have with “general words”. From the point of view of teaching translation, Kussmaul (1995: 109) remarks that “polysemy is the normal type of relation between a lexical item and its meanings. This is one of the first facts we should make our students aware of…” It was decided to capitalise on metaphorical extension and to encourage students to learn common-core polysemic forms used with different denotations in technical contexts (i.e. the general words in Dudley-Evans and StJohn’s second category), a way of providing the gentlest possible introduction to technical vocabulary. Pre-testing using DIY texts containing such general words confirmed that Willss’ remarks about lack of knowledge applied to my predominantly Italian learners too, in that, although students knew the forms in their common-core meaning, not only was there a lack of knowledge of the specific meaning of these items in the technical context, but also in some cases a lack of knowledge of the semantic notions underlying them, for it was difficult for a noticeable number of students to supply equivalences in their mother tongue.
2. A Learner-centred Methodology
The need for the polysemic items having thus been verified, we set about the process of acquiring them. This would be done by discovery methods, and in groups, since autonomy in choosing which items were worth learning and the empathy developed between learners (Dickinson 1987: 26) would reduce possible inhibition stemming from unfamiliarity with technical contexts. It was recognised that such methods would not permit the identification of all generally applicable polysemes in an exhaustive sense; but it was hoped that, if a small number could be discovered, the experience would raise students’ awareness of the phenomenon and stimulate them to look for further examples as a strategy at a later date. It was also hoped the learning process would prove educational as well as instructional, in terms of raising the learners’ consciousness of some common technical processes and components. The British National Corpus, as a large general corpus containing both common and specialist texts, provided the means for learners to verify whether lexical items originally selected by learners were in fact polysemic, and if so, the extent of the variation of their denotation in different technical contexts.
In an application of Candlin’s (1984,1987)learner-centred, negotiable task-based “process” syllabus,  the learners’ liberty extended as far as the freedom to interpret the aim of learning semi-technical vocabulary in terms of following general words other than polysemes, to select the uses they thought most worth learning, to include non-technical uses if they felt they were useful, to judge between the use of various monolingual and bilingual dictionaries and the corpus, and to decide for themselves how to present their research findings in written and oral form and what learning activities to design for other learners to help them learn the same items. The intention was not only, as has already been mentioned, to potentiate learning by attending to “affective” wants as well as cognitive needs, but also to allow for genuine experience of the different ways in which dictionaries and corpora can answer learning translators’ and interpreters’ vocabulary needs.
2.1 Pedagogic Procedure
The procedure adopted was, briefly, as follows.
1. Using extracts from DIY manuals, the teacher outlined types of words thought likely to turn up in a range of fields, together with general principles for the learning and retention of vocabulary (Gairns & Redman 1986), principally that vocabulary was more likely to be retained after some sort of controlled cognitive processing (Shiffrin & Schneider 1977).
2. Small groups of learners photocopied a target-language text or downloaded one from the Web. Operating instructions or technical product descriptions were the suggested genres since they are likely to turn up in the work of technical translators.
3. Learners identified a set of words in their text they thought might be “polysemic general technical lexical items”.
4. Learners made concordances of these items on the BNC, identifying which occurrences were “technical”, and verifying whether their word was indeed polysemic.
5. Students selected technical occurrences and read as much of the text as much of the concordance text as was supplied, seeking to understand it as fully as possible.
6. Students attempted, with the aid of the “source” function, and the native-speaker teacher, to identify the field of each text.
7. Selections of findings were arranged in booklets, including learning activities designed by the group to enable other learners to learn their items.
8. A selection of the findings was presented orally in plenary session.
It was explained to learners that stages 4 to 8, in involving cognitive processing, were intended to help them remember the items they worked on.
The results were variable in quality and divergent in focus, reflecting on the one hand the varied aims, learning style preferences, aptitudes , commitment and presentation skills of the learners, and on the other hand, the range of aspects of lexicographical skills experienced. I will present general results first, then give selected examples of what individual groups of learners presented. First of all, this list shows words or phrases verified as general polysemic technical items through their occurrence across different fields in the BNC.
3.1 General technical words
Table 1 Confirmed technical polysemes
(Confirmed in corpus, thus at least 2 technical meanings)
It is interesting to note that many of these forms form part of the common core of forms that every learner of intermediate level upwards is familiar with, and that many of them are not Romance language cognates, and hence without checking on the corpus they may not have been easily conceived of by learners as having their technical application.
3.2 Lexicographical Aspects
In his discussion of text analysis and the use of dictionaries, Kussmaul (1995: 124) states:
In order to be able to extract the maximum of information from definitions and examples in dictionaries, students should have some knowledge of the basic concepts of lexicography, which include structural semantics with notions such as synonymy, hyponymy, polysemy, hononymy, collocation, connotation and distinctive features. Looking up the meanings of words and the equivalents in dictionaries are only part of the truth, however. In many cases, we will not find the truth in dictionaries at all. We will then need techniques of analysing meaning in context.
The potential of polysemy projects in enhancing translating sub-skills is shown by the presence of many of the skills mentioned by Kussmaul in the following list of processes of translation or preparation for interpreting which were mentioned as useful aspects of the projects during semi-structured interviews administered to a chance sample of six learners who did polysemy projects in 2000.
• acquisition of basic generally-applicable technical terms
• understanding variation of meaning according to context
• building of lexical fields around one topic
• comparing appropriacy of entries in different monolingual dictionaries
• comparing appropriacy of entries in monolingual and bilingual dictionaries
Going beyond the dictionary to the context:
• comparing corpora and dictionaries as sources of information about language
• acquiring basic knowledge about technical processes
In addition, I observed the students experience the following processes:
• comparing entries under a single lemma to choose the one most appropriate for the context
• understanding term- formation through compounding
• understanding the metaphorical link between common and technical uses.
3.3 Polysemy projects: examples
The following extract from a student booklet presenting concordances and definitions is an example of how students grouped polysemic examples to gain awareness both of the underlying concept and the different fields in which it was applied. The source text was about bicycle mechanics.
(NOUN, COUNTABLE & ADJECTIVE)
Meanings in (source text) context:
a) In general, flat metal pieces as components of a bicycle’s mechanics.
If you don’t know which type of chain you are dealing with, clean the side plates and check the brand names [...] In particular, the metal disks belonging to the gear shifters of a bicycle. Gear shifters in general have the function to regulate a bicycle’s range of speed or power. The plates a shifter is composed of are defined as outer plate, ratchet plate and back plate, depending on their position and form (Milson 1995).
PRINTING: A sheet of metal which is carved or specially treated with chemicals so that it can be used to print texts or pictures.
Lithography a printing process based on the principle of the natural aversion of water to grease. The photographically prepared printing plate when being made is treated chemically so that the image will accept ink and reject water (BNC 1995: GOO 1959)
PHOTOGRAPHY: The format of a thin sheet of glass, plastic etc. that is covered with a layer of sensitised emulsions which react to the light and on which an image can be formed.
Half-plate pictures, if they are well cropped before being sent out, are equally acceptable. Pictures for television should be 10’’ x 8’’ matt, preferably in colour. For those publications which use colour, transparencies can usually be accepted from 33mm upwards, but it may help the magazine if the plate sizes are agreed before the pictures are taken. ((BNC 1995: EX6 1070).
GEOLOGY: A tectonic plate is large piece of the earth’s surface, perhaps as large as a continent, which moves very slowly.
Its size and its stability have come about largely because, unlike the other oceans, the Pacific is made essentially of one tectonic plate—one of the eight plates that cover the surface of the earth [...] There are essentially two processes at work, one to be seen at the western edge of the Pacific and the other towards the east.
In the west, the Pacific Plate meets two other major plates, the Eurasian and the Indo-Australian. The collision is slow, mighty and spectacular.” ((BNC 1995: CJD 504, 522)
One of the components of MACHINERY that constitutes its base.
“For most cuts the work is steadied against the foot plate which can be angled to
bring it into contact with the workpiece; because the teeth of the sawblade tend to
become worn close to the foot plate, the plate can be unclamped and moved
forwards to bring a fresh section of the blade into use ((BNC 1995: A16 971).
3.4 Term Formation
The importance of understanding term formation is noted by Sager (1981: 98), who notes that “specialized communication can be made more effective if terms are formed according to certain prevailing patterns which have a predictive value”. Like most of the students, this group thought it worthwhile to note compounds formed with their polyseme, which gave them practice of compounding principles, a skill that is useful when translators are forced by lack of dictionary information to make a reasonable guess of a term translation on the base of known words. It is notable that the group indicated their awareness that the noun “plate” functions as an adjective in some compounds . Naturally the students also became aware of how the corpus can help decide which compound combinations are legitimate. Here is an example extract from a student project.
dental plate ~ A dental plate is a piece of plastic which is shaped to fit inside a
person’s mouth and which a set of false teeth is attached to.
plate glass* ~ Thick glass made in large, flat pieces, which is used especially to
make large windows and doors.
plate tectonics* ~ (GEOL.) The way that large pieces of the earth’s surface move
slowly around and interact with each other
home plate [sing./BASEBALL] ~ The place where you stand to hit the ball and
the last place the player who is running must
touch in order to get a point.
plate armor [MIL.]* ~ Defensive armor of strong metallic plates for protecting
ships or fortifications against artillery.
~ [HIST.] Defensive armor for the person made of
*[the noun “plate” functioning as an adjective]
In a further example of work on term-formation, here is a list of terms generated by combinations of the adjective -operated, which, while not, in fact polysemic, was identified by the students as particularly “generative”.
• dynamo-operated lamps of Russian bicycles
• battery-operated syringe pump
• pen-operated computer
• one-person-operated trains
• power-operated sliding doors
• key-operated lockable handle
• computer-operated typesetting
• mouse-operated graphical display
• hand-operated plunger
• electrically-operated signal box /mechanically operated boxes
• touch-operated computer terminals
• coin-operated solarium
• foot-operated pedal bins
• power cable-operated brakes
• baby-operated tape player
• robot-operated assembly line
• toe-operated rubber pedals
• thumb-operated rocker switch
• treadle-operated Singer machine
• driver-operated buses
• laser-operated radar
• handle-operated brakes
• platform doors were an air-operated sliding type
3.5 Lexical fields
In an interesting divergence from the main aim of acquiring generally applicable technical vocabulary, several groups, including one composed entirely of interpreters, found the corpus useful to generate specialist lexical fields in various domains. Each field was built up as a result of checking in dictionaries the meanings of words found in a single concordance entry for one polysemes. Here is an example.
LAMPS AND LIGHT
1) A decorative security light with a clear or frosted glass globe, a traditional lantern light in black or white, and a floodlight security kit with a powerful 500watt halogen bulb. All will welcome you or your guests, but startle intruders. All lamps have a built-in passive infra red sensor which reacts to invisible thermal radiation emitted by people or other sources of heat.Any movement detected by the sensor is signalled to the electronic switching circuit which automatically turns on the lights.Using a pre-set timer, the lights can be set to stay on for a period from 25 seconds to 12 minutes. (BNC 1995: A16 745)
Bulb: the glass part of an electric lamp, which gives out light when electricity passes through it.
Electronic switching circuit: a closed system of wires which can start or stop operating by means of a switch, turn on or off.
Floodlight: very powerful lamp that is used outside to light public buildings, sports grounds and other places at night.
Frosted glass: a frosted glass has had its surface roughened so that you cannot see through it clearly.
Halogen: any of the group of elements fluorine, chlorine, bromine, iodine and astatine ( together form group VII of the periodic table) which are typically gaseous or volatile, strongly electronegative non-metals readily forming binary compound with metals.
(built-in passive) infra red: infra red equipment detects infra red radiation.
Lantern: a lamp in a metal frame with glass sides and with a handle on top so you can carry it.
Pre-set: if a piece of equipment is preset, its controls have been set in advance of the time you want it to work.
Sensor: an instrument which reacts to certain physical conditions or impressions such as light or heat, and which is used to provide information.
(invisible) thermal radiation: divergence from a central point of energy, especially heat, that comes from a particular source.
Timer: a device that measures time, it is especially part of a machine and causes it to start or stop working at specific times.
The figure of nine defined items per concordance entry here is about average; though the maximum number was 17. This group built 38 such fields; useful work, but which could have been carried a stage further, since several of their fields were related, but had not been grouped together. Nevertheless, the inference is that by this method learners can build up large amounts of vocabulary prior to interpreting assignments.
Many of the students found that there were certain forms where the link between common and technical uses was often more than polysemic, and could be qualified as metaphorical or even analogous, as with “fuel starvation”. Students were intrigued that arms, feet and even nipples fulfilled remarkably similar functions in the human body and in machines. Again, this could be said to give insight into the process of term formation, and help learners select from alternatives in dictionary entries with more confidence.
3.6 Acceptability to Learners
As regards the method’s acceptability to learners, all the interviewees found the autonomous style of learning enjoyable, and could envisage using the BNC in their future work as translators or interpreters, in that, unlike the dictionary, it provided reasonably long, and thus more authentic contexts to enable them to more reliably score hits as regards choosing natural and appropriate target language forms. Reservations expressed were that working translators would be hard pushed to find the time to consult the BNC, especially if synonyms were more readily available from other sources, and that using it to check the field-reference of a term would be easier if the exact field of each entry, rather than just the large and heterogeneous “domain”, was indicated on the screen to make such information as accessible as it is in dictionaries.
As already mentioned, following polysemes through the BNC gave learners experience of a number of lexicographical processes that need to be part of their model of working as translators and interpreters. In addition, the polysemy projects provided learning translators and interpreters with useful experience of the scope of a general corpus, and a comfortable transition from common core language to technical concepts and vocabulary. Experiencing practical vocabulary through purposeful concordancing on a general corpus with the facility to skip between fields, or between examples in the same field, seems a motivating way of providing input for possible acquisition of base technical concepts and appropriate target language lexis. The learning of the basic polysemes themselves might not only give learners a basic survival kit of practical lexical items, but also enable them, first, when translating from L2, to deduce meanings when familiar forms are encountered in new fields, creating an intuitive awareness of the underlying principles of term formation which allow native speakers to often arrive at reasonably accurate definitions of technical neologisms . Secondly, in translating towards L2, awareness of these basic forms should contribute to building a translator’s ability to select appropriate entries from the range displayed under a lemma,  but might even contribute to forming the “genius” of translators’ hunches about appropriate target language forms, enabling them , when no precedents are available, to themselves create or re-create plausible L2 neologisms.
On the methodological side, the learner-centred project-type approach succeeded in motivating learners to learn technical language, which in a more formal approach might have been perceived as daunting or inhibiting. The learners’ positive evaluation of the projects confirms that language acquisition can be perceived as a useful component of translating and interpreting courses if it is learnt through motivating, co-operative methods involving groupwork.
A reservation that should be expressed is that the subjects in this study have mainly learnt their L2 through a mixture of communicative methods and the study of literary extracts, and have received no previous technical education of any kind. One can thus conclude that, for learners with similar backgrounds, corpus-based polysemy projects could serve as a useful introductory transition to technical translation and interpretation before passing on to more specialist register analysis.
Finally, a reflection on awareness-raising as an approach to learning translation. I must admit that, each time I set this polysemy project in motion, I have the nagging doubt that educating translators about technical vocabulary by the use of corpora of short extracts, as opposed to the detailed study of longer texts, is merely skipping over the surface. But each time the project has run its course, I feel refreshed by the idea that it is only by moving over the surface of a labyrinth, that one sees the course of the thread of connections that can speed one across it.
Baker, M.(1988).“Sub-technical vocabulary and the ESP teacher: An analysis of some rhetorical items in medical journal articles”, Reading in a foreign language 4,2: 91-105.
Bernardini Silvia (2000). “I corpora nella didattica della traduzione: dall’addestramento alla formazione”, I Corpora nella Didattica della Traduzione / Corpus Use and Learning to Translate. Proceedings of the International Workshop, Bertinoro 14-15 November 1997, Bernardini Silvia / Zanettin Federico, eds. Bologna: CLUEB. 81-106. (English version, “A ‘trainee’ translator’s perspective on corpora”, on line, http://www.sslmit.unibo.it/cultpaps/paps.htm)
BNC (1995).British National Corpus Version 1.0 . Oxford: Oxford University Computing Services. [Accessible at ]http://info.ox.ac.uk/bnc/]
Candlin, C. (1984).“Syllabus design as a critical process”, General English Syllabus Design, ELT Documents 118, Brumfit, C.J., ed.. Oxford: Pergamon &the British Council.
Candlin, C. (1987). “Towards task-based language learning”, Lancaster Practical Papers in ELT vol. 7, Candlin C. / Murphy D. eds. London: Prentice-Hall.
Dickinson, Leslie (1987). Self-instruction in Language Learning. Cambridge: Cambridge University Press.
Dudley-Evans, Tony / St.John, Maggie Jo (1998). Developments in ESP: a Multi-disciplinary Approach. Cambridge: Cambridge University Press.
Duff, Alan (1981). The Third Language. Oxford: Pergamon.
Gairns, Ruth / Redman, Stuart (1986). Working with Words. Cambridge: Cambridge University Press.
Gavioli,Laura / Zanettin, Federico (2000). “I corpora bilingui nell’apprendimento della traduzione. Riflessioni su un’esperienza pedagogica”, I Corpora nella Didattica della Traduzione / Corpus Use and Learning to Translate. Bernardini Silvia / Zanettin Federico, eds. Bologna: CLUEB. 61-80 (English version, “Comparable corpora and translation: a pedagogic perspective”, on line, http://www.sslmit.unibo.it/cultpaps/paps.htm)
Gile, Daniel (1995). Basic Concepts and Models for Interpreter and Translator Training. Amsterdam & Philadelphia: Benjamins.
Kussmaul, Paul (1995). Training the Translator. Amsterdam & Philadelphia: Benjamins.
Lyons, J. (1995). Linguistic Semantics: an Introduction. Cambridge: Cambridge University Press.
Malmkjaer, Kirsten (1998). “Introduction: translation and language teaching”, Translation and Language Teaching, Language Teaching and Translation, Malmkjaer, Kirsten, ed. Manchester: St.Jerome. 1-11.
Martin, A.V. (1976). “Teaching academic vocabulary to foreign graduate students ”, Tesol Quarterly 10,1:91-9.
Milson, Fred (1995). The Bike Book. Sparkford, Somerset: Haynes.
Nation, I.S.P. (1990). Teaching and Learning Vocabulary. Boston, Ma: Heinle & Heinle.
Sager, J.C. (1981). “Approaches to terminology and the teaching of terminology”, Fachsprache 3,3:98-106.
Shiffrin, R.M. & Schneider, W. (1977). “Controlled and automatic human information processing, II: perceptual learning, automatic attending, and a general theory”, Psychological Review 84:127-90.
Weatherby, Joanna (1998). “Teaching translation into L2, a TT-oriented approach”, Translation and Language Teaching, Language Teaching and Translation, Malmkjaer, Kirsten, ed. Manchester: St.Jerome. 21-37.
Wilss, Wolfram (1996). Knowledge and Skills in Translator Behavior. Amsterdam & Philadelphia:Benjamins.
Process approaches are educational rather than instructional; the differences in these approaches as regards the learning of translation are highlighted by Bernardini (2000).
They might thus avoid mistakes caused by automatically selecting from a lemma the entries that are cognate with the source-language.