Sektionsleiter | Section Chair: Bernd Spillner (Duisburg)
|
Internet-Zeitschrift für Kulturwissenschaften | 17. Nr. | Februar 2010 |
| Sektion 6.4. | Internationale Fachkommunikation in Wirtschaft und Recht Sektionsleiter | Section Chair: Bernd Spillner (Duisburg) |
||
Quantitative Data Presentation in German Research Articles in Economics
Irena Vassileva (Bernau bei Berlin, Germany)
Email: vassileva_irena@gmx.net
1. Introduction
Harwood (1995:48) describes the ‘clash of academic cultures’ at the time when German scientists – refugees from the Nazi regime emigrated to the USA - as follows: “The émigrés were struck by their American hosts’ preoccupation with method and measurement while the Americans were amazed by their guests’ predilection for theorizing on a grand scale”. The Americans were an academic “community that was predominantly experimental”. A similar claim is made by Galtung (1985) in his seminal paper on intellectual styles of scientific communication. These observation are, however, relatively outdated and based on general impressions rather than concrete investigations.
Clyne (1993:9) compares quantitatively and qualitatively English and German academic texts in linguistics and comes to the conclusion that: „Some cultures, for instance the English-speaking ones, are more formally oriented than others. [...] A content orientation is related to the cultural idealisation of science and to the authority of the scholar or the intellectual.“(1)
2. Aim of the study
In order to check the above claims in view of modern scientific argumentation practices the present study aims to look at the ways and means of quantitative data presentation in German economics articles with special attention to the expression of approximated quantities. The functional-communicative values of the elicited vague expressions are discussed in terms of Channell’s (1994) and Dönninghaus’ (2005) classifications. Further on, the investigation focuses on the linguistic means of realisation of approximations. The presentation of statistical data is evaluated in terms of layout and quantity with particular emphasis on data treatment and discussion (upgrading / downgrading the importance of data, ignoring, approximating, covering for lack of specific information, etc.). The results are then compared to Channell’s (1994) observations on English economics articles and German linguistics articles.
3. Data
The investigation is based on 10 research articles in economics written by native speakers of German and published in the leading peer-reviewed journal “Zfbf : Schmalenbachs Zeitschrift für betriebswirtschaftliche Forschung”. The choice of the field of economics was dictated by the assumptions that, first, it should provide texts presenting statistical information and therefore containing both vague and precisely expressed quantities. Second, it is still understandable to me (I hope!). The topics of the articles cover a wide variety of areas.
4. Theoretical framework
Following Channell (1994:20), “An expression or word is vague if:
Thus, approximating quantities in scientific writing is usually a matter of author choice between exact and vague expressions without changing drastically the underlying proposition. Variations are observed along the axis of choice in immediate context, not along the axis of general context, since in its general context science is believed to describe and explain the world with precision and accuracy, where the latter “would seem to be desirable particularly for the presentation of quantities” (Channell 1994:96). Among other things, the choice depends on the purpose of the particular text, the intended audience, the concrete communicative situation of text production and consumption, the media used, as well as “the extent of the speaker’s knowledge of what is being approximated” (Channell 1980:472).
In terms of communicative function, “[…] vague expressions of quantity are used to present statistical data in a way that favours the argument of the author, but still conforms to academic conventions of truthfulness in the presentation of data” (Channell 1994:111), thus authors have to try to maintain the “delicate balance between factual reporting and persuasive writing” (ibid., p. 111). Therefore, much of the research on expressions of quantity in scientific writing relies heavily on Grice’s maxims for explanations of the observed phenomena, especially on the maxim of quantity: “Do not make your contribution more informative than is required” (Grice 1975:45). How much is, however, required? Is the amount of information language-, culture-, discipline-specific? These are questions still to be answered.
Another point that should be made clear is that the investigation deals with the notion of vagueness only and leaves the discussion of precision outside its scope. I shall just mention here for reasons of clarity that continuous, indiscrete measurements are often imprecise for epistemological reasons, while “Discontinuous, discrete measurements allow precise and certain statements” (Pinkal 1985:261, for a detailed discussion see also Dönninghaus 2005).
5. Methodology
The investigation follows roughly Channell’s (1994) classification of the ways and means of expressing vague quantities. She identifies the following types:
In this study the use of round numbers was not statistically investigated for technical reasons, namely it would have required detailed interviews with authors in order to find out whether each particular occurrence was a ‘real’ round number or a rounded one. Besides, not all groups and subgroups from Channell`s classification were observed in the corpus. Thus, within Group A the corpus showed no occurrences of the combination ‘n or so’ and just a few – of ‘n or m’. Group C was represented only by two subcategories – ‘vague quantifier + countable noun’ and ‘adverbs of frequency’. Groups D and E had no representatives in the corpus either. The missing categories after Channell are obviously reserved for types of communication other than scientific since they include items belonging primarily to colloquial language.
However, I observed two additional means of expressing vague quantities in scientific writing, namely (1) nominal phrases of the type ‘(adverb) + (adjective) + noun’ where at least one component refers semantically to quantities, and (2) verb phrases where the semantics of the verb contains the element ‘comparing quantities’. Thus, the following groups were eventually taken into consideration for quantitative and qualitative comparison between English and German academic writing:
6. Results and discussion
6.1. General observations
Authors make use of a very large amount of statistical data presented both in the form of (traditional) tables containing either numbers or percents and other (modern) means of graphic presentations (pie charts, graphs, etc.). That is, they make use of the latest technological achievements in graphic design available nowadays in contrast to linguists, for instance, who still stick to the well-familiar table (Vassileva forthcoming).
The numbers included in the tables and in the texts are not only whole (rounded) numbers and frequently contain more than one digit after the decimal point. Rounding in general is avoided, which means that more detailed numeric data is relevant for the communicative needs of economists.
More often than not authors tend to repeat the numbers given in the tables throughout the text itself and then comment on them; English linguistics writers, for instance, rely on the reader consulting the data in the tables and offer more detailed comments in the text.
This is most probably one of the reasons why there are only 25 occurrences of vague quantity expressions in the texts in 108 pages, which makes a frequency of 0.12 items per page. On the other hand, in German linguistics texts the frequency is 0.36 per page, i.e. three times more (Vassileva forthcoming). In English linguistics articles, on the other hand, the frequency is as high as 0.66 items per page. One could find a tentative explanation of this phenomenon in Hall & Hall’s (1990) description of American and German cultures in general: both are considered to be low-context cultures where “the mass of communication is vested in the explicit code” (p. 6), where detailed information is needed in the process of communication; “[…], Americans, Germans, and French all need information, facts, and figures. However, each defines and uses data in different ways. Americans are primarily interested in ‘the bottom line’. Germans want lots of examples with figures” (ibid., p. 104). On the other hand, we seem to be observing differences between the two fields of science, economics being more precise and relying more and more on mathematical methods and linguistics being interested primarily in overall tendencies and dispositions.
The quantitative distribution of the linguistic means for the expression of vague quantities found in the corpus is presented in Table 1:
About / around / approximately + number |
7 |
Partial specifier + number |
8 |
(Adverb) + (adjective) + noun |
7 |
Verb phrase |
2 |
Vague quantities + noun |
1 |
Altogether |
25 |
Table 1
The data show that the most frequent linguistic means is a nominal phrase, which, in a way, confirms the image of the language of science as nominalized. In contrast to the language of linguistics there are no occurrences of adverbs of frequency and the use of verb phrases is brought to a minimum. As will be seen later in this paper, this is often due to the fact that vague expressions are frequently accompanied by the respective real values, usually in brackets. In any case, the present pilot study cannot provide an in-depth explanation of this phenomenon due to its narrow focus and limited data.
6.2. ‘About / around / approximately + number’
These approximations consist of an approximator, an exemplar number and an optional measure noun. Channell’s (1994:44) tests with native speakers of English have shown that number approximations have three basic features: (1) They designate intervals of numbers which are symmetrically distributed about the exemplar number; (2) The length of the interval increases with the increase of the size of the exemplar number and, (3), the nature of the items affects the length of the interval.
The three most frequent approximators of this type in the German language are ungefähr, circa (zirka) and etwa. Only two of them, however, are represented in the present corpus, namely etwa and circa, where the second one appears only in its Latin spelling. Almost all occurrences are with whole round exemplar numbers referring to the number of people interviewed for a given study:
Ex. (1) Zur Befragung von Kunden eines Internet Service Providers werden ca. 12,000 Kunden ...
Ex. (2) Das größte analysierte Team, [...], besaß etwa 300 Mitglieder.
In two cases etwa is observed with percents, but both refer to comparison of exact data presented in tables:
Ex. (3) Beide Studien beziehen sich jedoch auf Situationen, in denen die Flatrate etwa 13% teuerer ist als ...
In contrast to linguistics, where the exemplar number is most frequently a simple fraction, there are no such cases in the present corpus.
Channell (1994:83) maintains that “[…] in a natural category such as numbers, not all members are equivalent. Certain members of the category serve as reference points to which other members are related” and those may be approximations. Further on she notes that: “Reference point numbers are a product of the base number system being used (in Britain, base 10). But some reference points in the number system are culturally determined” (ibid., p. 94). It seems, then, that economists prefer to round following the reference points of the decimal system in order to make their data more accessible and easier to process by other economists (and possibly by a wider audience).
6.3 Partial specifier + number’Partial specifiers “specify upper or lower limits for quantities on the number continuum” (Channell 1994:62) and are the most frequent in the corpus. Almost all of the partial specifiers (actually, there is only one exception) denote lower limits:
Ex. (4) Einer der befragten Teams, […], bestand aus über 120 TM aus etwa 50 Ländern.
Ex. (5) Da jedoch bei Stichprobenumfängen mit mehr als 400 Beobachtungen...
In contrast to the first group there are cases of use of fractions as exemplar numbers, but the only observable fraction is half:
Ex. (6) Tatsächlich haben von den 39 Going Private-Unternehmen, [...], 22 und damit mehr als die Hälfte während der Zeit ihrer Börsennotierung keinen Bedarf an zusätzlichem Eigenkapital.
Ex. (7) Folgerichtig antworten mit neun mehr als die Hälfte der 17 ...
Ex. (8) Knapp die Hälfte davon (24,3 Prozent) haben...
Besides, as the examples demonstrate, the vague expression is actually specified in terms of quantities even within the same sentence, so that it is difficult to determine whether such formulations should really be treated as cases of vagueness.
The partial specifiers denoting upper limits account for only two cases and refer to time spans:
Ex. (9) […] die Entsandten mit höchstens 64 Monaten Aufenthaltsdauer […]
One could thus tentatively say that German economists tend to look for regularities, for features that predominate rather than focus on exceptions.
6.4 (Adverb) + (adjective) + noun’This is the largest group of linguistic means for expressing vague quantities. In terms of Quirk’s semantic classification, practically all adjectives are stative and permanent, i.e. they denote unchangeable characteristics of the presented data:
Ex. (10) […] statistically significant difference […]
Ex. (11) Die Computerinstruktionen dagegen haben einen deutlichen Überhang […] [The computer instructions on their part have a clear surplus]
All adjectives are non-restrictive and non-inherent; they describe features of quantities which may be found in many other entities:
Ex. (12) […] GP’s mean was slightly higher than MO’s […[
Most of the adjectives are, as expected, gradable and often occur in the comparative or in the superlative:
Ex. (13) […] hinsichtlich vieler Kriterien sogar signifikant bessere Werte.
Ex. (14) […] der Anteil [ist] signifikant niedriger […] [the share is significantly lower]
One also observes the use of absolute (non-gradable) evaluative adjectives denoting a very high degree:
Ex. (15) In Übereinstimmung mit den Ergebnissen der Hit Rates zeigen sich auch beim MAE eindeutig überlegene Werte für [...]
Many of the adjectives of frequency are further intensified by degree or emphasizing adverbs:
Ex. (16) Die deutlich besten Werte […]
It seems that scholars believe to be more convincing if they further emphasize their quantitative results.
Sometimes the vague quantity is expressed by a noun:
6.5 Verb phrase’Ex. (17) […] keine eindeutige Überlegenheit festgestellt werden kann.
Ex. (18) Bestätigt wird darüber hinaus die Annahme, dass es in der Mehrheit der Fälle [...]
As Table 1 shows, ‘verbs of frequency’ account for only 2 cases of the vague expressions of quantity, whereby both occurrences are of verbs denoting high frequency:
6.6 Vague quantities + noun’Ex. (19) Die Faktorreliabilitäten übersteigen in allen Fällen 0,6 […]
Ex. (20) Während bei der Einbeziehung sämtlicher Unterersuchungseinheiten noch die lineare Regressionsvariante dominiert, [...]
The only vague quantifier found in the corpus is viel:
Ex. (21) Deshalb gab es in vielen der von uns befragten Teams noch Kritik […]
In the language of linguistics, as well as in the above example, such quantifiers appear before nouns denoting persons who have been observed or interviewed for the purposes of a particular study. They make it possible for the author to refer anonymously and vaguely to his/her subjects of investigation, which is usually done for ethical reasons.
7. ConclusionsAn earlier longitudinal study of the language and structure of economics articles in German (Vassileva 2004) demonstrated that at the beginning of the 20th century scholars made very little use of data. Later on, in the middle of the century and especially from the 70-ies on statistical and mathematical methods became gradually indispensable for research in economics.
The results of the present investigation show a definite shift among German economists towards more empirical, statistically supported research. This fact demonstrates once again the far-reaching consequences of the process of globalisation – the Anglo-American rhetorical patterns of academic discourse presentation have irrevocably permeated even those fields of the humanities and social sciences which are most dependent on a particular language and culture, in this case a language of an academic area with long-standing and powerful traditions.
Second, scientific language does not make use of the whole variety of linguistic means for expressing vague quantities established in earlier studies (mainly Channel’s), but also makes recourse to certain verb and noun phrases. Obviously, a large part of vague quantity expressions are reserved for spoken / colloquial language.
Third, the use of vague quantity expressions in scholarly writing is primarily dictated by the following considerations (after Channell 1994):
This is a small-scale preliminary investigation both as far as the quantity of data is concerned and in terms of the scientific fields covered. Further studies are necessary based on larger (possibly electronic) corpora and including various areas of academic writing, as well as psycholinguistic experiments in order to be able to elicit the culture- and language-specific strategies scholars employ in the presentation of statistical data and explain cross-cultural misunderstandings that may ensue from established differences. The results of such contrastive investigations may serve as a basis for the creation of teaching materials for academic writing courses.
Corpus
References:
Anmerkungen:
1 All translations in the article are mine.
6.4. Internationale Fachkommunikation in Wirtschaft und Recht
Sektionsgruppen| Section Groups| Groupes de sections
|
Inhalt | Table of Contents | Contenu 17 Nr. |  |
Webmeister: Branko Andric last change: 2010-03-12