On the reconstruction of an individualised language

Note-taking in consecutive interpreting.


Introduction
Consecutive interpreting is typically used for press conferences, after-dinner speeches and similar occasions.The statements to be interpreted can be as long as 20 minutes.As the capacity of the human memory is insufficient to provide a consecutive of longer statements, the interpreters make notes to support their memory and thus to facilitate the rendition in the target language.
A typical notation is given in appendix A. What appears to be a modern form of hieroglyphics is, in fact, an excerpt from the 1992 consecutive interpreting finals (English into German) held at the Institute of Translating and Interpreting at the University of Heidelberg.Readers unfamiliar with the applied notation system will undoubtedly find the notation completely incomprehensible.A natural language transcript will not be of much help (see appendix B).And what is more, even readers who are familiar with the underlying system will hardly be able to understand the extract without knowing the source text and communicative context that gave rise to its production (see appendix C).
What then is the linguistic nature of a notation?Note-taking is commonly regarded as some kind of supporting technique, developed by practitioners for practitioners to help them retrieve part of their source text understanding from memory.In order to fulfil this function, note-taking aims at keeping the processing effort as low as possible.Three basic principles can be identified that are largely undisputed in specialist literature: -Economy: to minimise the processing effort any notation should be as scarce and brief as possible.-Instantaneous seizability: the strain on the memory can be effectively relieved only if the interpreter can read the notes at a glance.-Individuality: note-taking is not governed by any obligatory rules or regulations.Generally speaking, anything that supports its function or that is subjectively felt to do so is admissible.
What is most controversially discussed, however, is the relationship between notation and natural languages.The controversy, which is rooted in the translingual dimension of the consecutive interpreting task, revolves around the alleged language-independent nature of notation.The discussion is dominated by two prominent (yet not always clearly distinguished) claims.The first claim makes a semantic point and argues that successful interpreting involves a deep and comprehensive understanding of the source text and thus requires a notation that is able to grasp the source text's meaning as a deverbalised entity (Seleskovitch 1988;Seleskovitch & Lederer 1989).(Note that this is a text-as-object approach, which incorporates the assumption of one text having one meaning.) The second claim concerns the nature of note-taking signs and emphasises the need for an approach that makes them as language-neutral as possible (Matyssek 1989).The underlying idea is that the interpreter's detachment from the source text surface structures can only be achieved with the help of a notation system that stays clear of any of the languages involved.In addition, it is also seen as a considerable advantage if notation systems are applicable to all of the interpreter's working languages (see Rozan 1956:9).In practice, however, notation systems show clear evidence of source or target language influences -without detrimental effects on the interpreter's performance.Ilg (1980) is the only author to stress not only the importance of a deep analysis and understanding of the source text's sense but to emphasise also the major role played in interpreting by the source text's expression side.He points to the fact that especially on the level of international gatherings, a main field for interpreting, an extremely codified, ritualised and formalised language prevails which, in many cases, has to be retained or at least taken into consideration.He, therefore, vehemently rejects Seleskovitch's "mépris du mot" (1980:118).
The traditional debate about the language-independence of notation leaves many questions unanswered.What is needed is a paradigm shift from the still prevailing view of notation as a language-independent 'technique' towards a thorough linguistic understanding of the issues involved.Against the background of a cognitive theory of language and discourse, this paper will argue that note-taking can (and should be) understood as an interpretingspecific discourse process based on an interpreting-specific notation language.

Notation as language
From a cognitive perspective, a language can be understood as a reservoir of linguistic means of expression -from sounds to words and phrases to grammatical structures -that are designed to help speakers create utterances that convey their meaning intentions and hearers to construct an interpretation of those utterances and intentions.Notation systems likewise provide means of expression suitable for producing notation utterances and for conveying utterance meanings.A notation system is actually a notation language offering a surprisingly wide range of notation signs with lexical, syntactic and pragmatic values.So far only few attempts have been undertaken to explore this dimension, viz.Kirchhoff (1979), Allioni (1989) and Kalina (1998:183).
Notation signs make deliberate use of natural languages and are shaped by various reduction, adaptation and iconisation processes.It would be completely uneconomical not to have recourse to natural languages and, instead, to invent and learn new, arbitrary means of expression.At the same time, it can be shown that a notation language has its own notation-specific means and develops its own specific structures and characteristics, which, in line with functional requirements, are quite distinct from those of natural languages.
Notation languages are designed to minimise processing costs on different dimensions: they should be maximally suited for fast, economical and effective note-taking, easy to learn, and in compliance with the preferences and strengths of the individual interpreter.For this reason, notation languages typically make use of the most varied self-created or borrowed means of expression, and they do so by openly choosing them in terms of the advantages they have to offer.It is here that an explanation should be sought for the fact that studies have found the mixed character of notation language to be a common practice among professional interpreters (see Ilg & Lambert 1996:80,88) as well as in student training (see Ahrens 2001).
There follows a short overview and preliminary linguistic categorisation of some of the notation signs developed by Matyssek (1989) and used in our notation example (see appendices A and B).

Word meanings
A principle function of natural language lexemes is to enable users to activate conceptual structures in their minds that are associated with the lexemes by way of their denotational meanings.The same holds for lexical notation signs.This is quite apparent in cases where the notation lexeme is represented by a lexeme from either the source or the target language, e.g.irony A striking feature of notation language is the extensive use of iconic symbols to represent notation lexemes.Such symbols are simplified, stylised and shortened pictographic signs whose denotational meaning is recoverable from their iconic form.Following Lyons (1977:155), we will regard iconicity as "a more specific kind of motivation".Iconic notation lexemes are language-neutral and they have a high degree of associativeness, memorisability and seizability.Here are a few examples of iconic motivation: Source notation lexeme denotational meaning motivation C13 world globe with equatorial line C14 development spiral moving forward C12 knowledge derived from belief (Christian cross) as is conviction The knowledge example, which is derived from belief, shows that the motivation of notation signs may have a cultural background, e.g. the occidental Christian tradition.
The principle of iconicity may also be exploited by joining conceptually related notation lexemes with similar pictographic representations to form pictographic groups at the lexical level.In that way it is possible to cover comprehensive lexical fields, e.g.

Source notation lexeme denotational meaning motivation
to wish one egg in a basket C14 to will/want several eggs in a basket to demand many eggs in a basket Many notation lexemes are characterised by an unusually high degree of vagueness and much of their interpretation relies on context information (compare words such as the colloquial expression thing or the adjective good).While vagueness is a well-known linguistic phenomenon, in notation language this trait is even more widespread.It clearly helps to cover a wide range of potential meanings while at the same time keeping the number of signs that have to be learned limited and, thus, manageable.Examples include the following: Source notation lexeme denotational meaning C11 to say, to speak, to express, to declare, to make known, to convey, to point out, etc. and (if the context makes it possible as is the case here) to predict C1 to enjoy, to laugh, to smile, funny, happy, ironic, etc.

Word formation and inflection
Notation language deploys a variety of word formation devices from different language sources to ensure economical and effective processing.
Various natural language sources and shorthand signs are used in abbreviation, e.g.Bound notational lexemes may be attached to any notation sign, be it an abbreviation (e.g.Partnerschaft), a pictographic symbol (e.g.Gemeinschaft) or an initial.The use of affixation opens up highly creative possibilities.Thus, Wissen and -ung are combined to form Wissen(svermittl)ung .This shows that, for functional reasons, notation language does not imitate source structures (unless absolutely necessary, e.g. to avoid confusion), but develops its own inner structures.
Word formation mechanisms such as compounding and blending are also frequently used, e.g.

Wirtschaft
Entwicklung Wirtschaftsentwicklung Wirtschaft Politik Wirtschaftspolitik Finally, notation language also makes available a number of inflectional morphemes.The context is generally clear enough to make gender classifi-cation superfluous, even though the possibility of adding an -e for the feminine subclass theoretically exists.Number is a morphological category which may be either recovered from the context or marked explicitly.To mark the plural a superscripts is attached to the notation sign as a bound morpheme.Adding a double superscriptss expresses the concept many (or marks some kind of stress, e.g.source text: there are complaints -loud complaints → notation text: complain ss [C17]).Tense affixation is also often discarded (as in the international community will be meeting [C8]) although this may potentially lead to severe distortions in the interpreting process.To avoid such distortions, the graphic angle-signs and may be used as a bound morpheme.If it points to the right, it indicates the future, while the past can be expressed by an opening to the left; mode can be indicated by turning the angle's horizontal line into a wavy line.As far as the grammatical subclass of case is concerned, only the expressive genitive form is made explicit in the form of a slash (as in content of agenda [C11], poor nations of world [C13], interest of rich [C18]).

Semantic relations
At sentence level, semantic relations between phrases and clauses are indicated by structuring the notation text in a vertical, indented and terraced way, which is called "verticalisme/décalage" (Rozan 1956:13, 19-22) or "tiering" (Ilg & Lambert 1996:82); dashes in the margin function as full stops, thus delimiting sentences or coherent units.
An impressive example is the prepositional phrase in C14 (see appendices A and B), a stronger and more effective partnership between environment and development.The concept of between is visually expressed by putting the sign for partnership between those for environment and development.What is more, the co-ordination of elements is indicated by putting one on top of the other, e.g. the adjectives strong and effective.For a full visual effect, these adjectives are put below the sign for partnership so as to point out clearly which noun is modified by the adjectives.
In the same way, clauses can also be co-ordinated by putting one on top of the other.Subordination is indicated by putting the clause down to a lower level and indenting it, as is the case with relative clauses and thatclauses [C2, C12, C17].In addition, relative clauses are sometimes put in brackets, while that-clauses may be set off by a comma [C12, C17].Indentation is not necessary in the case of conjunctional clauses as these are clearly marked by putting the conjuncts in the margin [C3, C5, C9, C12].
Taken together, these graphic means serve to visualise the text's semantic structure which helps to grasp its main points immediately.Because of this structuring function, the layout must be seen as a notational means of expression.

Pragmatic markers
Illocutionary functions are explicitly expressed in the margin so as to signal to the interpreter whether to formulate an assertion (unmarked), a question (question mark in the margin, often the Spanish question mark) or an exclamation (exclamation mark in the margin).
Similarly, stylistic effects, ironic or humorous remarks, political insults or other noteworthy connotations are marked in the margin to draw the interpreter's special attention to particular points where special care might have to be taken or where a more moderate formulation might be appropriate, especially with a view to culture-specific target audience sensitiveness.

Notation as discourse
When compared with the abundant richness of natural languages, a notation language seems to be poor and restricted, even deficient.How can such a limited resource be of any help at all in the interpreting process?We propose to deal with this question in the broader context of discourse analysis.
From a cognitive point of view, discourse comprehension involves the creation of a mental representation of what the text is about in the recipient's mind.Depending on the depth of processing, this mental representation is built up at different levels from the textual meaning of words and phrases to the propositional and illocutionary meaning of individual utterances and a holistic mental model of the text's overall content (see van Dijk & Kintsch 1983, Johnson-Laird 1983).Linguistic knowledge, world knowledge and knowledge about the immediate communicative situation are equally important in this process.They are activated and deployed in a complex and strategic combination of data-driven (bottom up) and expectation-driven (top down) processes (see Brown & Yule 1983, Schnotz 1994).Following this text-as-process approach, a written (or spoken) piece of discourse is experienced as a text if (and only if) the recipient is able to construct for it a coherent mental representation of its communicated meaning.For an application of this approach in the context of simultaneous interpreting see Kohn & Kalina (1996).
The most striking characteristic of human communication is the miraculously wide gap between the little that is explicitly said and the wealth of what is implicitly meant -and understood (Grice 1975).How can this be accounted for by a theory of discourse comprehension?According to the principle of relevance (Sperber & Wilson 1986/1995), each utterance is geared towards achieving a maximum contextual effect for a minimum of processing cost, i.e. towards balancing cost (i.e.processing effort) and reward (i.e. gain in information).The following example from Blakemore (1992) illustrates the main points: A) Did you enjoy your holiday B) The beaches were crowded and the hotel full of bugs In compliance with the principle of relevance, speaker A assumes that speaker B's reply is maximally relevant in the context in which it was uttered.This enables speaker A to activate whatever linguistic or nonlinguistic knowledge she might need to produce the required contextual effects, i.e. inferences that help her understand what is meant.On a first processing dimension, this involves explicating, that is fleshing out the "propositional skeleton" specified by speaker B's utterance: she thus has to assign reference to "the beaches" (which beaches?) and "the hotel" (which hotel?), to disambiguate the polysemous lexeme "bugs" (insects or spying devices?) and to enrich "crowded" (with whom?).The fully explicated propositional meaning is then taken to a second processing dimension where the implicated meaning ("No, I didn't enjoy it at all.") is inferred.
The important point here is that the processes of explicating and implicating reach out far beyond the explicitly marked interpretation clues the utterance provides.A diversified range of linguistic and non-linguistic knowledge is strategically activated and deployed to satisfy the essential requirement of the utterance's maximal relevance.Explicating and implicating processes thus enable us to recover what is meant by an utterance.At the same time, however, the intrinsically subjective nature of utterance comprehension becomes apparent: depending on the background knowledge available to the individual recipient, an utterance will be understood in different ways or not at all.
From a consecutive interpreting perspective, the relevance approach is highly adequate and enlightening, since note-taking is by its very function and nature geared towards minimising processing cost.The point here is not just to reduce the number of notes taken to a minimum, but to find the optimal balance between noted (i.e.explicit) and memorised (i.e.implicit) information.The goal for the interpreter is to be able to retrieve a maximum of information from memory using a minimum of notation without straining her cognitive resources in the process.As a consequence, the lack of explicitness can assume surprising proportions in notation utterances/texts; explicating/implicating processes are usually required to reach out much further than in natural language texts.
It is, for instance, quite common for single notation signs to be used to represent a wide range of facts, which in the source text require a whole passage or more.The single sign for greeting (°) serves to illustrate this point.It may be used as a bottom-up cue to represent all the detailed source text information relating to the welcoming of the Queen and her entourage by the President of the Federal Republic of Germany plus the reason for her visit and perhaps the mentioning of the Queen's last visit to Germany or that of the German President to England.To recover all this, the interpreter has to resort to extensive disambiguation, enrichment and reference assignment processes.
In order to complete the explicatures and implicatures of the notoriously reduced and underdetermined notation utterances, ellipsed and unexpressed content has to be recovered from memory to a great extent.It is in particular the typical notational concentration on the more meaningful and informative lexical elements which brings about the characteristically condensed notational lexemes, phrases or clauses that have to be completed and enriched in the process of notation text reception.
In our notation example (see appendix) this point is illustrated by the reduction of the prepositional phrase to establish an agenda for the environment and development to the phrase environment development [C9], or by the clause but one thing of which I am fairly certain is this, which is simplified to but [I] know [C12].In the same way, a notation usually omits referents which are obvious from the co-text (people [C20]), semantically weak verbs that only serve stylistic purposes or can be inferred from the immediate context (to be [C1, C2, C17, C18], to see [C14], to meet [C8]), or collocates that vary from language to language and can automatically be retrieved by the interpreter from her knowledge about the target language (source text: make decisions → notation text: decisions [C4]).
One might tend to believe that the typical condensation and concentration on the lexical elements in notation text production does not necessarily imply that there has to be completion and development of these condensed forms in notation text reception, but that it should be sufficient to recover, on the basis of the reduced notation text and memory of the source text, the global content and implicit intentions of that source text.Contrary to the ordinary conference participant, however, the interpreter is expected to strive for utmost completeness (see Seleskovitch 1988:46-7).This means that in notation text reception it is not only the global content and a few interesting details which count, but also the facts and details at the level of local coherence.These facts are expressed by the speaker in the propositional form of the intended explicatures.As a result, the recovery of the explicatures in notation text reception is a central part of the interpreting task and is probably done in a more conscious way than in natural text reception.
Notation is all about keeping the right balance between explicit and implicit.While notation texts in general are condensed to a degree that make them "odd" in comparison to typical natural language texts, this trait does by no means constitute a necessary condition for what constitutes a notation text.Balancing out the explicit and the implicit is largely influenced by a number of external and internal factors determined by the immediate interpreting situation and the interpreter's individual predispositions and skills.
External factors include the structure and density of the text, speed of delivery, sound quality, ambient noise, the number of listeners, etc. Internal factors include general and specialised background knowledge, general language and terminological knowledge, intelligence, the ability to concentrate and to remember, tiredness, motivation, etc.Consequently, while the call for a full and deep understanding of the text is legitimate, one has to bear in mind that in interpreting the level of understanding may vary as a function of the above parameters.
This becomes clear from looking at interpreting in more practical terms.It can be shown to be less cost intensive, at times, to jot down information by means of a couple of short, simple and easily recognizable notation signs than to keep them present in memory and to have to reconstruct or reactivate them later on, even if the information is well known.In other cases it may be totally uneconomical to note down fully grasped information in detail, since this would draw the attention to the act of note-taking and take it away from the essential act of listening to and analysing the source text.In yet another case, e.g.towards the end of a tiring conference day, detailed note-taking may help a well informed interpreter to take a slight break and relax her overburdened cognitive apparatus.
Detailed note-taking often runs the risk of closely following the source text structures and, thus, of interfering with target text production.One reason why the interpreter stays close to the surface structure of the source text may be because the expressions used in the source language are particularly important, e.g. in contract negotiations.Another reason could be that the interpreter failed to understand the utterance properly.As a result, the mental representation formed in the interpreter's mind is only a lowerlevel representation of the text's propositions rather than a fully developed mental model.In this case, it is inevitable that the notes taken by the interpreter reflect the structure and language of the source text.It follows that minimalistic notation is not possible on all occasions.Nevertheless, it is still the ideal notation to be aimed at and the one which most noticeably reflects the specific character of notation texts.
Successful note-taking is greatly facilitated by detailed knowledge of the subject matter in question.This goes to show that the interpreter's preparation of the subject matter and the related terminology, which is a prerequisite for successful comprehension, has an immediate effect on the notation text.Similarly, the provision of slides, PowerPoint presentations or brief outlines of a speech as well as the repeated presentation of information in the framework of successive interpreting appointments may affect the balance between explicitly noted and memorised information.The not unfamiliar case that a fully formulated written text has been supplied to the interpreter beforehand and is therefore known to her in the course of her interpreting task will most certainly be reflected in her notation text in such a way that most of the relevant information remains implicit.
In order to fully understand why and how an interpreter is able to benefit from a notation text, its specific role and purpose need to be taken into account.The primary purpose of a notation text is to bring back to its own producer, the memory of the source text, i.e. to help her reconstruct and reactivate the textual meaning (mental model) that was constructed in the process of source text comprehension.In this sense, a notation text is an ancillary text which depends for its raison d'être on the source text.Comprehension of a notation text thus largely depends on the interpreter's ability (a) to build up and memorise a mental model of the source text and (b) to create and use a notation text that optimally supports retrieval of the mental model (or portions of it) she managed to memorise.
In addition, a notation text is produced for immediate and exclusive use in the limited temporal framework of a concrete interpreting task.And what is more, it is produced by, addressed to, and interpreted by one and the same person, i.e. the interpreter.A notation text need not -and usually does not -contain the linguistic clues deployed in natural language texts to enable their comprehension for third party listeners, or at a later point in time outside the immediate production context.
The highly implicit and idiosyncratic nature of notation texts is thus functionally related to the special processing conditions of consecutive interpreting, and is as such a feature the interpreter is able to cope with.A notation can be described as a text, which is produced by the interpreter and whose meaning is recovered by her in the process of notation text reception via standard cognitive comprehension and processing mechanisms but which, at the same time, differs from natural language texts in view of its function, special processing conditions and, as a consequence, its balance of explicit and implicit information."It is an irony -though not a very funny one -that CITIES should itself be an endangered species.If we are to save the Convention, then this meeting must make a number of decisions about the future of CITIES, for we cannot go on the way we are doing at present.Ladies and Gentlemen: Less than three months from now, the international community will be meeting in Brazil to try to establish an agenda for the environment and development as we enter the next century.I cannot predict to you the content of that

B -Natural language transcript of the notation
Communicative context: The following is a speech by the Executive Director of the United Nations Environment Programme, Dr Mostafa Tolba, to the 8th Meeting of the Parties of CITIES.CITIES stands for the Convention on International Trade in Endangered Species of Wild Fauna and Flora.This Convention, worked out under the responsibility of UNEP, has been in force since the mid-seventies.